Tony wrote, "... I have uploaded?one of my old schematics ..."

Tony, I am wondering, what does the "SRC" stand for, in the filenames?

Thanks,
Andy

John Woodgate wrote, referring to the SDD model:

Could it be 'translated', if it's not too big?


If one had a copy of the SDD model in question, I am sure it could be translated to Spice syntax with some unknown amount of effort, however be aware a lot of ADS SDD models are written in Verilog and Ptolemy. Others are in a more obvious syntax closer to spice. The SDD is a container for various supported syntax’s.

Ray

The Master of Science Thesis uploaded here (/g/LTspice/files/Temp/Electrical%20Pulsing%20of%20a%20Laser%20Diode%20for%20Usage%20in%20Fluorescence%20Microscopy.pdf) contains the simulation with LTspice of a step recovery diode circuit.

I have found out following detail regarding setting local and global temperatures, read LTspice code below.
Main point is, that you cannot pass the temperature variable in a function to the subcircuit and as a value expression.
Has anyone made a similiar experience or explanation for this detail?

Here, the relevant code, simply copy into a simulation file.
<<
; Case 1: Direct usage of TEMP variable as parameter (e.g., voltage): => overwriting works.
.subckt SUB1 A B
V1 A B {TEMP}
.backanno
.ends

; Case 2: TEMP variable in function and usage of the function in a component parameter (e.g., voltage): => overwriting works.
.subckt SUB2 A B
.func f(x) {TEMP}
V1 A B {f(1)}
.backanno
.ends

; Case 3: Usage of TEMP variable in VALUE expression of controlled source: => overwriting works.
.subckt SUB3 A B
E1 A B VALUE={TEMP}
.backanno
.ends

; Case 4: TEMP variable in function and usage of the function in VALUE expression of a controlled source: => overwriting DOES NOT work.
.subckt SUB4 A B
.func f(x) {TEMP}
E1 A B VALUE={f(1)}
.backanno
.ends

.op

X1A S0 0 SUB1
X1B S1 0 SUB1 TEMP=50 ; ==> works
X2 S2 0 SUB2 TEMP=50 ? ; ==> works
X3 S3 0 SUB3 TEMP=50 ? ; ==> works
X4 S4 0 SUB4 TEMP=50 ? ; ==> DOES NOT work
>>

I have just uploaded the SPICE model of the MA44769-287T Step Recovery Diode as found in MACOM's website.

It is here: /g/LTspice/files/Temp/MA44769-287T_SPICE_MODEL.zip

?

Tony's schematics have the opposite problem of Helmut's from 2006.? Whereas the fonts in Helmut's are tiny, Tony makes his schematics with text that overlaps other text, making many of his schematics difficult to read.? That's because Tony set his LTspice to a different system font size, which allows him to place his text too close together without realizing it.

Go into the Control Panel (Tools > Settings > Drafting Options) and change the Font Properties > Size to something smaller.

I think the default size is 28, so that is what the majority of LTspice users have.? Whatever you change it to, remember that the number is saved and becomes your new default font size.

Andy

开云体育

On 23/05/2024 13:12, Andy I wrote:

Tony, you forgot to include your symbol file "diode-steprecovery.asy".

Oops. Yeah, I do that too. ? Fixed.

--
Regards,
Tony

Tony, you forgot to include your symbol file "diode-steprecovery.asy".

Andy

abid wrote, "Is the whole circuit SRD? It's a bit confusing," referring to step_recovery_diode.asc from 2006.

No, I don't think so.? Only about half of it is.

Also, the text is rather small, but you can fix that by starting LTspice with the "-FixUpSchematicFonts" command-line switch.? This schematic pre-dates a change in LTspice's handling of font sizes, which makes them look so small today.? Use "-FixUpSchematicFonts" to update the font sizes.

On that schematic, look for the comment text "Diode with parasitics".? The components below that text are:

• Cse
• Cp
• Lp
• D1
• Cpad1
• Cpad2

Those six components represent the SRD.? Well, technically the first four are the SRD, and Cpad1 and Cpad2 are probably the PCB pads to which the SRD connects.

All the other stuff to the left of Cpad1 and to the right of Cpad2 are external devices to which the SRD is connected.? It looks like Tshunt is a shorted quarter-wave trap filter, and T1 and T2 just represent traces on the PCB.? Rs is the source impedance and RL is the load.

Helmut provided several URLs as comments on the schematic.? It might be worth following those, to see where he got this schematic.? Since it is nearly two decades old, you might need to use the "Wayback Machine" () to retrieve some of them if they are not on the 'net anymore (but hopefully they are).

However - Tony is more of an expert on this, so I recommend following his recommendations.

Andy

开云体育

On 23/05/2024 10:37, abid wrote:

On Wed, May 22, 2024 at 08:39 PM, Andy I wrote:

I'm sorry, I forgot to include the link to the 2006 schematic file uploaded by Helmut Sennewald, with his step recovery diode.? It is here:

Files > z_yahoo > Examples > Educational > step_recovery_diode.asc
/g/LTspice/files/z_yahoo/Examples/Educational/step_recovery_diode.asc

His schematic notes say you must use the "Alternate Solver", but it seems to work with the "Normal Solver" too.

Is the whole circuit SRD? It's a bit confusing.?

It's just the diode itself that's the SRD. I have uploaded one of my old schematics that uses the SRD in shunt mode, which is often more useful than the series mode. It also doesn't use an LC matching circuit, which makes a wideband design possible. Normally, SRD comb generators and multipliers tend to be quite narrowband (I/P PRF).

--
Regards,
Tony

On Wed, May 22, 2024 at 08:39 PM, Andy I wrote:

I'm sorry, I forgot to include the link to the 2006 schematic file uploaded by Helmut Sennewald, with his step recovery diode.? It is here:

Files > z_yahoo > Examples > Educational > step_recovery_diode.asc
/g/LTspice/files/z_yahoo/Examples/Educational/step_recovery_diode.asc

His schematic notes say you must use the "Alternate Solver", but it seems to work with the "Normal Solver" too.

Is the whole circuit SRD? It's a bit confusing.?

开云体育

Could it be 'translated', if it's not too big?

The SDD element is a
‘Symbolically Defined Device’ that is used in Agilent ADS simulator.

The ADS syntax for the SDD element is definitely not Spice compatible.

Ray Anderson 

I'm sorry, I forgot to include the link to the 2006 schematic file uploaded by Helmut Sennewald, with his step recovery diode.? It is here:

Files > z_yahoo > Examples > Educational > step_recovery_diode.asc
/g/LTspice/files/z_yahoo/Examples/Educational/step_recovery_diode.asc

His schematic notes say you must use the "Alternate Solver", but it seems to work with the "Normal Solver" too.

Andy

There is another schematic, from 2006, uploaded by Helmut Sennewald, which also claims to use a step recovery diode.? The diode model in that circuit is an "ordinary" SPICE diode using a .MODEL statement:


I can not confirm any of these models.? I just present them as evidence of what other LTspice users did.

Andy

FYI -

There are some step recovery diode models and simulation schematics in our group's Files, from seven years ago.? I refer to these two directories, although beware that some of the uploaded files in them are inconsistent or missing things:

Files > z_yahoo > Files sorted by message number > msg_102029
/g/LTspice/files/z_yahoo/Files%20sorted%20by%20message%20number/msg_102029

Files > z_yahoo > Files sorted by message number > msg_102029
/g/LTspice/files/z_yahoo/Files%20sorted%20by%20message%20number/msg_102890

The .ZIP file in the second of these might be complete.? But it simulates VERY SLOWLY.? I do not know what its original uploader was attempting to do.? Simulating a 200 GHz oscillator for 0.55 seconds seems to me just a little bit odd, and the output file would be ginormous.? Attempting to run that simulation now, I see math errors - voltage values of -1.#IND, which is an error code from the Windows math library.? Some of the file descriptions suggest 400 MHz, but comments on the schematics say 200 GHz, so I would regard them with a great deal of skepticism.

But you can see the SPICE model they used for their step recovery diode.? If nothing else, it might be worth studying how it works.? Or maybe not.? Their models are Subcircuits, visible on the schematic sheets.


Andy

An image of a spice model for a SkyWorks SRD has been extracted from CVB1031 SRD data sheet and place in the “SRD model and topologies” photo album:
/g/LTspice/album?id=295221

i haven’t posted a link to the data sheet as it is on an external site and such links are frowned upon in this group.

-Ray

I have never tried to simulate a SRD using ADS. Apparently there is an Agilent appnote on the topic but I haven’t been able to track it down.

The last time I worked with step recovery diodes was back in the early 1980’s when I was at California Microwave. I wasn’t involved in the SRD frequency multiplier design then, ?but as I recall we had a proprietary simulator written in-house that was used for SRD simulations.

-Ray

On Wed, May 22, 2024 at 05:46 PM, Ray Anderson wrote:

The SDD element is a
‘Symbolically Defined Device’ that is used in Agilent ADS simulator.

Do you now how to design the SDD in ADS based on the equation and then use that SRD model in other circuits? I could try with the ADS then. Thanks.

The SDD element is a
‘Symbolically Defined Device’ that is used in Agilent ADS simulator.

The ADS syntax for the SDD element is definitely not Spice compatible.

Ray Anderson

Andy I wrote, "If these are indeed the same elements with the same values in both figures, then it suggests that the "SDD" component represents only these parts of Figure 3: Cr, Rf, Cf, Vo, and the switch.? The switch might be key to making it model the step-recovery diode."

?

Thanks. I also figured that out. But I was wondering how can I design the SDD component in LTSpice? Or should I design the equivalent circuit? I also need to incorporate the charge equation somehow.