Signal generators - Agilent 33220A or Tek AFG3021B?

[Philip Pemberton-2]

Hi guys,
I'm in the market for a signal generator -- I'm trying to get a PLL to
lock onto an RF (low hundreds of kHz) carrier, then use a phase detector
to decode the ~12.5bps phase-modulated data that's riding on top of the
carrier.

Yes, I'm making a Radio 4 Timecode decoder. Though IIRC the BBC spec
calls it "LF Radio-Data".

Catch: the Hameg signal generator I'm using now doesn't do any form of
modulation besides some quick-and-nasty AM which I've been using to
tweak the filters. 1024-point 12-bit arbitrary waves is about as
advanced as it goes, which isn't going to get me a 198k sine with PM data.

So I'm looking at a new signal generator -- the Agilent 33220A and Tek
AFG3021B are on my shortlist and have basically identical specs (aside
from the Tek's bigger display). I'm leaning towards the 33220 because
it's smaller than the Tek (desk space is at a premium), and it's going
to be tethered to a PC most of the time anyway so the large LCD is
pretty irrelevant. It's also about £200 +VAT cheaper...

Now before I plunk down ~£1k +VAT on one of these boxes... has anyone
actually used one of these boxes? Any comments or criticisms?

Most of what I'll be using it for is <20MHz or so RF testing, bandwidth
tests / frequency response of amplifiers, that sort of thing. I'm also
open to suggestions of similar equipment from other manufacturers...

Thanks,
Phil.

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Vasile Surducan-2]

Well, you may get the Agilent at about $400-500 which is good, but if
you need some modulations up to 20Mhz carrier, why don't you consider
a DDS like AD9851 ?
It's definitely cheaper if you have some spare time... and you can
design yourself a quite good signal generator (forget it only if you
need SFDR bigger than 60-70dB and you don't have a good spectrum
analyzer)

Vasile

On 7/11/08, Philip Pemberton <piclist@...> wrote:

> Hi guys,
> I'm in the market for a signal generator -- I'm trying to get a PLL to
> lock onto an RF (low hundreds of kHz) carrier, then use a phase detector
> to decode the ~12.5bps phase-modulated data that's riding on top of the
> carrier.
>
> Yes, I'm making a Radio 4 Timecode decoder. Though IIRC the BBC spec
> calls it "LF Radio-Data".
>
> Catch: the Hameg signal generator I'm using now doesn't do any form of
> modulation besides some quick-and-nasty AM which I've been using to
> tweak the filters. 1024-point 12-bit arbitrary waves is about as
> advanced as it goes, which isn't going to get me a 198k sine with PM data.
>
> So I'm looking at a new signal generator -- the Agilent 33220A and Tek
> AFG3021B are on my shortlist and have basically identical specs (aside
> from the Tek's bigger display). I'm leaning towards the 33220 because
> it's smaller than the Tek (desk space is at a premium), and it's going
> to be tethered to a PC most of the time anyway so the large LCD is
> pretty irrelevant. It's also about £200 +VAT cheaper...
>
> Now before I plunk down ~£1k +VAT on one of these boxes... has anyone
> actually used one of these boxes? Any comments or criticisms?
>
> Most of what I'll be using it for is <20MHz or so RF testing, bandwidth
> tests / frequency response of amplifiers, that sort of thing. I'm also
> open to suggestions of similar equipment from other manufacturers...
>
> Thanks,
> Phil.
>
> --
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
>

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Alan B. Pearce]

>Most of what I'll be using it for is <20MHz or so RF testing, bandwidth
>tests / frequency response of amplifiers, that sort of thing. I'm also
>open to suggestions of similar equipment from other manufacturers...

Keithley and Thurlbey Thandar are both in this market. IIRC you have already
bought a Tek DPO from TTi, so give them another shout. Otherwise Keithley
have a range of RF generation gear, they appear to have been quietly getting
stronger in this area, as well as having some nifty little booklets and CDs
on doing RF measurements. I haven't been keeping tabs on precisely what they
have, but from glancing through the leaflets that come across my desk, they
would be worth a look.

Probably also worth looking at Yokogawa, they seem to be the main Japanese
supplier of good quality instruments.

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Philip Pemberton-2]

Vasile Surducan wrote:
> Well, you may get the Agilent at about $400-500 which is good, but if
> you need some modulations up to 20Mhz carrier, why don't you consider
> a DDS like AD9851 ?

No FM or phase modulation...

> It's definitely cheaper if you have some spare time... and you can
> design yourself a quite good signal generator (forget it only if you
> need SFDR bigger than 60-70dB and you don't have a good spectrum
> analyzer)

I am looking into using a couple of the more advanced ADI DDS chips to
generate a QPSK test waveform (and I can probably reprogram that chip to do
BPSK with a +/-22.5-degree shift at 198kHz).. It's a nice idea, but I wanted
to have a go with AM/FM receivers too - and neither the Hameg nor the ADI
chips can do that (easily). AM involves an SA612 multiplier; FM is a lot more
painful, especially with a DDS.

Thanks,
--
Phil.
piclist@...
http://www.philpem.me.uk/
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Dave Tweed]

Philip Pemberton wrote:

> Vasile Surducan wrote:
> > Well, you may get the Agilent at about $400-500 which is good, but if
> > you need some modulations up to 20Mhz carrier, why don't you consider
> > a DDS like AD9851 ?
>
> No FM or phase modulation...
>
> > It's definitely cheaper if you have some spare time... and you can
> > design yourself a quite good signal generator (forget it only if you
> > need SFDR bigger than 60-70dB and you don't have a good spectrum
> > analyzer)
>
> I am looking into using a couple of the more advanced ADI DDS chips to
> generate a QPSK test waveform (and I can probably reprogram that chip to
> do BPSK with a +/-22.5-degree shift at 198kHz).. It's a nice idea, but I
> wanted to have a go with AM/FM receivers too - and neither the Hameg nor
> the ADI chips can do that (easily). AM involves an SA612 multiplier; FM
> is a lot more painful, especially with a DDS.

Huh? One of us is very confused.

I thought most DDS chips has frequency and phase offset registers that made
FM and PM dead easy...

Even if your modulating signal is in the analog domain, you just digitize
it with a suitable ADC with a suitable sample rate, and feed it on through
to the DDS. What am I missing?

-- Dave Tweed
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Philip Pemberton-2]

Alan B. Pearce wrote:
> Keithley and Thurlbey Thandar are both in this market. IIRC you have already
> bought a Tek DPO from TTi, so give them another shout. Otherwise Keithley
> have a range of RF generation gear, they appear to have been quietly getting
> stronger in this area, as well as having some nifty little booklets and CDs
> on doing RF measurements.

For what it's worth, it's not specifically RF I'm looking for -- just the
modulation schemes.

For instance I'm after phase modulation (BPSK) for the Radio4 timecode
receiver. The plan being to get the PLL to lock onto a "perfect" signal, then
work on the RF front-end and filters later.

It's still a toss-up between the Tek and the Agilent though -- the Tek will go
to 25MHz and has a dual-channel version. The Agilent is better specced for
interfacing -- USB, LAN and GPIB as standard (the Tek just has USB; you only
get GPIB and LAN on the 100MHz and 200MHz versions).

The Thandar generators are nice, but the lack of any real modulation schemes
makes them little better than the Hameg. The sequence/loop mode on their
generators looks useful though, but the price/performance ratio seems a little
poor. I suppose I shouldn't really be comparing a DDS against a variable-clock
generator though.

I think I'm going to have a word with a few Tek distributors over the next few
days. Probably TTi first.

> Probably also worth looking at Yokogawa, they seem to be the main Japanese
> supplier of good quality instruments.

No price list on the website, which I take to mean "if you have to ask, you
can't afford it"... That's proven to be a fairly accurate assumption in the past.

Thanks,
--
Phil.
piclist@...
http://www.philpem.me.uk/
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Sean Breheny]

Dave is correct. ADI makes several DDS ICs which can very easily do
AM, FM, and PM simultaneously. They do AM using a mutiplying DAC which
is referenced off the output of another DAC. Some can also do FSK,
BPSK, and ASK directly (i.e., you configure the chip and then you can
just feed in a serial data stream on one pin and it modulates it).
Some also can automatically do frequency sweeps (i.e., let's say you
need to sweep frequency faster and smoother than you can do via the
control interface - you can program it to sweep linearly from freq A
to freq B over time interval T, and then trigger it and it will
execute it). They can pretty much do any type of signal generation you
want.

I've used the AD9852 with great results.

Sean


On Mon, Jul 14, 2008 at 4:55 PM, Dave Tweed <pic@...> wrote:

> Philip Pemberton wrote:
>> Vasile Surducan wrote:
>> > Well, you may get the Agilent at about $400-500 which is good, but if
>> > you need some modulations up to 20Mhz carrier, why don't you consider
>> > a DDS like AD9851 ?
>>
>> No FM or phase modulation...
>>
>> > It's definitely cheaper if you have some spare time... and you can
>> > design yourself a quite good signal generator (forget it only if you
>> > need SFDR bigger than 60-70dB and you don't have a good spectrum
>> > analyzer)
>>
>> I am looking into using a couple of the more advanced ADI DDS chips to
>> generate a QPSK test waveform (and I can probably reprogram that chip to
>> do BPSK with a +/-22.5-degree shift at 198kHz).. It's a nice idea, but I
>> wanted to have a go with AM/FM receivers too - and neither the Hameg nor
>> the ADI chips can do that (easily). AM involves an SA612 multiplier; FM
>> is a lot more painful, especially with a DDS.
>
> Huh? One of us is very confused.
>
> I thought most DDS chips has frequency and phase offset registers that made
> FM and PM dead easy...
>
> Even if your modulating signal is in the analog domain, you just digitize
> it with a suitable ADC with a suitable sample rate, and feed it on through
> to the DDS. What am I missing?
>
> -- Dave Tweed
> --
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
>

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Philip Pemberton-2]

Sean Breheny wrote:
> Dave is correct. ADI makes several DDS ICs which can very easily do
> AM, FM, and PM simultaneously. They do AM using a mutiplying DAC which
> is referenced off the output of another DAC.

Not exactly difficult then...

> Some also can automatically do frequency sweeps (i.e., let's say you
> need to sweep frequency faster and smoother than you can do via the
> control interface - you can program it to sweep linearly from freq A
> to freq B over time interval T, and then trigger it and it will
> execute it). They can pretty much do any type of signal generation you
> want.

Sounds like exactly what I need, only without the multi-£1,000 price tag...

I've just put in an order for a couple of AD9852s, and I've got some other ADI
DDS chips in my box to play with in the meantime.

AM + PSK could be quite handy - test without the modulation to get the PLL
stuff working, then add some audio to the PSK carrier to test the sideband
filter. Finally, add the RF front end and make that work...

Sounds like a plan, and a nice project to boot. It gives me a good reason to
finally learn about L-C filter design too...

Thanks,
--
Phil.
piclist@...
http://www.philpem.me.uk/

--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist