EEVblog Electronics Community Forum
Products => Other Equipment & Products => Topic started by: Rufus on July 17, 2011, 04:02:50 pm
-
You can view all details and specifications here http://www.tti-test.com/go/tg5011/index.htm (http://www.tti-test.com/go/tg5011/index.htm)
What you don't get there is the photo of the guts below.
Scanning from top left we have:-
Little PCB with usb host controller for the usb memory stick interface
Real mains switch below that
OEM PSU module, +5 and +/-15 outputs
Small fan, virtually silent
Main PCB containing amongst other stuff:-
NXP LPC2468 processor.
2 SRAMs which appear to be battery backed and I assume hold arb waveforms and settings
A Cyclone EP3C16 FPGA and a 4Mb SDRAM. AD9744 DAC.
The big heatsink bottom right looks like it is just a voltage regulator.
Haven't used it much yet but so far just a couple of niggles in pulse generator mode to complain about.
I have owned a TG1010 generator for years and have found it to be a reliable workhorse which was a good part of the reason for choosing this one.
-
thanx for the info rufus. can you identify for me section A,B and C? what they do, where their output going and where their input coming from. i assume section A is the signal and trigger output and they get their input from B which in turn from C? section C is the source signal (FPGA) giving output to section B and A and signal output to the BNC at the back? more detailed explanation will be more appreciated, thanx.
-
thanx for the info rufus. can you identify for me section A,B and C? what they do, where their output going and where their input coming from.
Not really. I didn't look in detail and it would be hard work figuring it out if I did look. The signal starts at the DAC top left of C. The rest will be mostly analog. Low pass filters, pulse rise/fall rate controls, level attenuators, signal offset addition and an output amplifier. There are possibly parts on the bottom of the PCB, can't see much and I didn't want to remove it. The thing is fully self adjusting for calibration so there is probably some circuitry in the area for that.
-
its ok. if my assumption is correct, and your about self calibration. no doubt its a quality generator deserves the £900 price point.
-
its ok. if my assumption is correct, and your about self calibration
If you are interested in the circuitry required search the web for Agilent arb generator service guides. You will find schematics for at least one of their older models and probably block diagrams for newer ones.
-
Pulse generator mode niggles have been fixed with a firmware update.
I am left with some questions about pulse generator mode on this type of signal generator and the performance of similar generators from other manufacturers.
The spec claims uHz frequency resolution and 10ns pulse delay and width resolution. Other generators also seem to claim 1uHz frequency resolution.
I don't see how that kind of DDS generator frequency resolution can easily be achieved in pulse generation mode (without accepting a lot of jitter). The TG5011 appears to align pulse edges with a 1GHz clock so you have 1ns period resolution and when setting pulse period rather than frequency 1ns resolution is all that is offered.
1ns period isn't 1uHz resolution at anything above 1kHz.
Anyone else have a generator with pulse mode and looked at how it performs at the limits?
-
(http://i.imgur.com/D26ML.jpg)
-
(http://i.imgur.com/D26ML.jpg)
Not sure why you posted that, however, it demonstrates a pulse mode niggle which hasn't been fixed.
The actual frequency output is the reciprocal of 1005ns. If you set pulse period you are only given 1ns resolution. Setting frequency gives (at this frequency) you about 8 digits which don't do anything and IMO incorrectly in the status area displays the set rather than actual output frequency.
A user interface which handles frequency dependant resolution nicely is quite difficult to do, I could live with it displaying the actual frequency in the status area.
My real question is do other similar generators with pulse mode have a different hardware scheme which actually gives uHz pulse frequency resolution.
-
Apologies about resurrecting and old post, I'm trying to decide between the TTi TG5011 and the Tabor WW5061 and the issue highlighted in this post puzzles me.
The TG5011 clocks the DAC at 125MS/s which is one sample every 8 nsecs which corresponds roughly to the rise time. How can the resolution of the pulse period then be down to 1nsec? The jitter is listed as 300 psecs for pulse (plus 100 ppm of period) or 500 psecs for square waves which is half a cycle of a 1GHz clock corresponding to the 1nsec resolution.
This seems to imply that there is a 1GHz clock which is divided by 8 for the 125MS/s DAC clock but the DAC clock can be shifted up and down by a few cycles of the 1GHz clock over a period. Given that the maximum pulse frequency is 12.5MHz the minimum period is 80nsecs or 10 DAC cycles the losing or gaining of up to 4 nsecs must be spread over some but not all the cycles. This all seems rather complicated - does it make sense?
The Tabor also has an 80 nsec minimum period but its resolution is 20 nsecs not 1 nsec which corresponds to 1 cycle of its (up to) 50MS/s variable clock.
Is a 1 nsec resolution, as apposed to a 20 nsec a major advantage ? The Tabor with its variable clock should have lower jitter.