# Need help putting 6Z4 into PSU Designer

#### sbelyo

##### Member
I need help figuring out the values for putting the 6Z4 / 84 into duncans psu designer.

The tube datasheet is here
http://tubes.mkdw.net/sheets/127/8/84.pdf

Parameters in detail

Name: 6Z4

Type: VT

Dres: 0 ohms is my guess

Vlaw: Unknown (The Va/Ia power law. This is the power ratio of Va to Ia)

Vfac: Unknown (Voltage factor, used to factor the rectifier current after other calculations have taken place.)

Vpiv: 1250 V

Ipks/Ipkr: Unknown

Can somebody help with the equations?

#### CarlyBoy

##### Member
I too would like to know how to calculate those parameters. I'd happily do the grunt work for some unpublished rectifiers if someone demonstrated how it's done.

#### Robert McLean

##### Member
I have calculated these parameters for 6X4

DRes = 53.6587
Vlaw = 1.755718
Vfac = 0.000421

Your can leave DRes out and use slightly different values for Vlaw and Vfac, but I find I can get a much more accurate fit to published data curves by putting DRes in.

The way I get the values for the 3 parameters is through regression using Solver in an Excel spread sheet. Get several values of current vs voltage from spec sheets. Assume there is a certain equation using those parameters that relates voltage to current. In this case there are two parts to the equation due to DRes
Vd = V - I*DRes
and I = Vfac * Vd ^ Vlaw.

where I is the current from the spec sheet,
V is the voltage from the spec sheet,
and Vd is the voltage seen by the imaginary diode after the current has passed through the imaginary resistance DRes.
Yes, I know, its recursive, ie circular, so you have to iterate a few time, I find 4 times is enough.

Then do the regression.

Here is the excel file I use for extracting diode model parameters for a bunch of different diode model types, ie Childs law, Perugini model, PSUD model etc. Put your spec sheet data into the work sheet "TestData" then just tab to the particular model worksheet you want, and use Solver to extract the parameters. Spice models are automatically generated. And the PSUD line is generated that you can just paste into your PSUD diodes text file.

I realize this description is brief, and it assumes you are familiar with Solver and regression in general, but otherwise it would need a big tutorial to explain all this.

And if you are wondering why the data in the spreadsheet isnt for 6X4, its because I cant seem to find my 6X4 data, I have no idea why.

EDIT
OOPS, I just noticed you want 6Z4, not 6X4. I'll do the 6Z4 later, I have to go do the dishes.

#### Attachments

• diodemodels.zip
42.7 KB · Views: 50

#### Robert McLean

##### Member
OK, I have processed the 84/6Z4 data and this is the resulting line for PSUD

6Z4, VT, 0.00499997661784358, 1.52269995965923, 0.000631197710625865, 1250, 99, 0.18

#### sbelyo

##### Member
Thanks...

I could never have figured that out. I was just going to build a PSU and wing it with the values. Now I can get real close.

Thanks again