ODD Comparison - OLD Agilent vs New Siglent

[4asifm]

Hi All,

I am looking for an oscilloscope - I am mostly interested in homebrew, ham and such.

I have an offer of a Agilent 54616B 500MHZ 2GSa/s (OLD) and Siglent 1202X-E (NEW) for about the same price. Once is a digitizing scope another a digital scope. I also understand that Agilent doesn't have the fancy FFT transforms and advanced math functions of the Siglent. But then higher sampling, single-shot bandwidth etc makes it very attractive. But the Agilent is old, discontinued and probably if it goes bust on me, I wouldn't be able to repair it or get it repaired.

Considering that this is going to be my only scope for the next 5 years, is the (OLD) Agilent a better choice over the (NEW) Siglent?

Thoughts?

This is my first post - please ignore any mistakes I made

[tautech]

If and I don't see why, you're chasing the 500 MHz BW get the old HPAK and take the punt it keeps going.
However these little SDS1202X-E won't test your back when needing to move it and be confident they can accurately  resolve signals at least 10% beyond their rated BW and offer great sensitivity.
One of my best sellers.

[Grandchuck]

Recommended reading:
https://www.eevblog.com/forum/testgear/analog-vs-digital-scope/msg3013596/#msg3013596

[Fungus]

In the long term I think the extra features of the Siglent  will be far more valuable than the extra bandwidth of the old Agilent.

[OwO]

I had already replied to your other thread when I saw this one, so here is my reply copied from there:

For RF work you want the maximum possible bandwidth. If you want to accurately see the waveform on a power amplifier you need to be able to see at least up to the 5th harmonic, preferably more. I would personally ignore any scope without a 50ohm input. You won't be using the math functions, because anything calculated will be horse shit since there is no such thing as measuring voltages and currents at RF; the voltages you see will bear little relation to reality and the only thing you should expect to see is the waveform shape (not even that if you don't get your probing technique right).

[Fungus]

I would personally ignore any scope without a 50ohm input.

Having 50 Ohms built-in at the push of a button is convenient but you can use terminators/adapters:

[Wuerstchenhund]

Hi All,

I am looking for an oscilloscope - I am mostly interested in homebrew, ham and such.

I have an offer of a Agilent 54616B 500MHZ 2GSa/s (OLD) and Siglent 1202X-E (NEW) for about the same price. Once is a digitizing scope another a digital scope.

Not really. Both are digitizing digital scopes, and because there are no non-digitizing digital scopes these scopes are just called 'digital scopes' or 'DSOs'.

I also understand that Agilent doesn't have the fancy FFT transforms and advanced math functions of the Siglent. But then higher sampling, single-shot bandwidth etc makes it very attractive. But the Agilent is old, discontinued and probably if it goes bust on me, I wouldn't be able to repair it or get it repaired.

Considering that this is going to be my only scope for the next 5 years, is the (OLD) Agilent a better choice over the (NEW) Siglent?

The old Agilent 54600 Series were great general purpose scopes back then, and they are generally very robust and reliable. The real problem with the 54616B is that this is not one of the fast MegaZoom equipped 54600 scopes which come with large memory, it's a conventional model which while offering 2GSa/s sample rate only has a measly sample memory of 5,000 points (i.e. 5kpts). Which means that the sample rate (and thereby the useable analog BW!) will quickly drop like a rock on anything except the shortest time base, and it means the scope will only be able to capture extremely short sequences at adequate sample rates.

It also lacks most of the features you will find in any bottom-of-the-barrel scope today, like color display, usable FFT (the 54616B can do FFT but only with 1kpt), serial decode and so on.

So no, I would not invest any money into a 54616B, even if you need the BW. Which doesn't look like you do, anyways. Besides that, a 54616B isn't really worth more than $200-$250.

So yes, get the Siglent, and if you can stretch the budget a bit more then consider getting a 4ch scope like the Rigol DS1054z or the Siglent SDS1104X-E because there are many situations where having more than just two channels is really helpful.

[OwO]

Well the issue is, if you use an external 50ohm terminator, the length of stub of the BNC connector plus the scope's internal routing will make a full circle around the smith chart in less than 1GHz, which means it'll look like a short circuit by 500MHz. SDS120X-E has a 18pF input, which if that is constant would put the -15dB point at 62MHz. Likely that 18pF is measured at 1MHz, which means it can be far worse at higher frequencies. I would say the signal fidelity is no good above 50MHz or so.

Seriously, 18pF is enormous. Any cheap MMIC amplifier can give you a good match up to many GHz which means parasitic capacitance in the range of femtofarads. If I *really* had to go for a scope without a 50ohm input, the first thing I would do is open it up and solder a 50ohm resistor on the input trace close to the first chip it connects to. Yes, it means you need proper probes but that can be easily DIY'd too. As long as you don't attempt any high voltage or "floating" probing a resistor divider + cheap MMIC amplifier will give you performance that beats any budget or low cost probes you can get.

[nctnico]

Well the issue is, if you use an external 50ohm terminator, the length of stub of the BNC connector plus the scope's internal routing will make a full circle around the smith chart in less than 1GHz, which means it'll look like a short circuit by 500MHz. SDS120X-E has a 18pF input, which if that is constant would put the -15dB point at 62MHz. Likely that 18pF is measured at 1MHz, which means it can be far worse at higher frequencies. I would say the signal fidelity is no good above 50MHz or so.

This problem doesn't exist. Oscilloscope inputs have some series inductance too. As a result you can easely measure well over 200MHz with an external 50 Ohm terminator. Try it.

[vk6zgo]

I had already replied to your other thread when I saw this one, so here is my reply copied from there:

For RF work you want the maximum possible bandwidth. If you want to accurately see the waveform on a power amplifier you need to be able to see at least up to the 5th harmonic, preferably more.

How so?
The waveform out of an RF PA is expected to be as close as possible to a clean sine wave (allowing for modulation, of course), & determining harmonic level from waveform distortion in the time domain is a pretty dicey way of doing things, in any case.

A Spectrum Analyser, a selective voltmeter, & worst case, a good quality receiver in the the possible harmonic frequency range, will all do a better job.

I would personally ignore any scope without a 50ohm input. You won't be using the math functions, because anything calculated will be horse shit since there is no such thing as measuring voltages and currents at RF; the voltages you see will bear little relation to reality and the only thing you should expect to see is the waveform shape (not even that if you don't get your probing technique right).

[OwO]

The waveform out of an RF PA is expected to be as close as possible to a clean sine wave (allowing for modulation, of course), & determining harmonic level from waveform distortion in the time domain is a pretty dicey way of doing things, in any case.

A Spectrum Analyser, a selective voltmeter, & worst case, a good quality receiver in the the possible harmonic frequency range, will all do a better job.

No, at these frequencies you will certainly go for class E/F PA designs, and in a class F PA the drain voltage waveform should closely approach a square wave. A spectrum analyzer won't do any good here because it can't tell you the relative phase between harmonics and therefore the waveform shape. Even in an ordinary class C PA you will want to verify the drain waveform symmetry (e.g. to make sure you aren't going to get close to the breakdown voltage limit).

[OwO]

This problem doesn't exist. Oscilloscope inputs have some series inductance too. As a result you can easely measure well over 200MHz with an external 50 Ohm terminator. Try it.

Only shunt inductance will cancel out shunt capacitance. Series inductance only makes it worse, hence why I said input capacitance is larger than stated at higher frequencies. It's simple to test, just measure the scope input (with 50ohm terminator attached) with a network analyzer. You will see return loss go to shit way before you reach 100 MHz. What *will* work however is an inline attenuator at the scope input.

[nctnico]

This problem doesn't exist. Oscilloscope inputs have some series inductance too. As a result you can easely measure well over 200MHz with an external 50 Ohm terminator. Try it.

Only shunt inductance will cancel out shunt capacitance. Series inductance only makes it worse, hence why I said input capacitance is larger than stated at higher frequencies. It's simple to test, just measure the scope input (with 50ohm terminator attached) with a network analyzer. You will see return loss go to shit way before you reach 100 MHz. What *will* work however is an inline attenuator at the scope input.

That is what I wrote: external 50 Ohm terminator.