DC Conditions for d3a

izzy wizzy · Joined: 26 Jun 2005 Posts: 53 Location: Folkestone, UK

When using the D3a, the DC conditions as quite a way off reality and the data sheet as far as I can make out. Is the model more suited to AC and not DC?

I get 190V on the anode while the sim suggests 162V. That is quite a difference. Any ideas?

cheers,

Stephen

dave slagle · Joined: 22 Jan 2005 Posts: 1357 Location: NYC

what is your screen voltage? When I adjust your sim to get 150V on the screen, it looks pretty darn close to the published curves. Maybe your tube is off the mark.

dave

reVintage · Joined: 07 Sep 2007 Posts: 336 Location: Sweden

Works like a charm for me when I adjust Us to the stipulated 150V.

izzy wizzy · Joined: 26 Jun 2005 Posts: 53 Location: Folkestone, UK

OK I did make an error in adjusting things but I still don't get the right screen current in the model compared to the datasheet. I will go back to the real circuit to look at it closer to see exactly what I've done.

In the model, if I make the anode R 5k6 to get 180V on the anode and make the screen R 47k to get 150V on g2, then yes, the anode current is what I would expect from the datasheet (about 13mA) but the g2 current is 2.3mA in the model while the datasheet suggests about 4mA.

Any ideas?

cheers,

Stephen

reVintage · Joined: 07 Sep 2007 Posts: 336 Location: Sweden

Hi Stephen,
As there allways is a certain spread among tubes, why not be pragmatic and use the SGR to finetune the working point with ref. to Ua and Ia. The way I see it, Spice should just be used to get guidelines.

sbench · Joined: 28 Jan 2005 Posts: 191

For D3a, the g2 current in the model is slightly lower than it ought to be. In addition, the effective G1-G2 mu was a little too low. The grid to anode capacitance was wrong. While I was at it, I added the same saturation limiting I put on the parts I recently added, which better models the clip points. G2 current isn't necessarily a very well controlled parameter; that's why a voltage divider is generally recommended over a simple series resistor to supply G2. The new model also works reasonably well in triode mode too (G2 connected to anode). The older model isn't as good triode connected.

I'll send the new file in the "updated tetrode.txt" thread.

Stephie <3

izzy wizzy · Joined: 26 Jun 2005 Posts: 53 Location: Folkestone, UK

Thanks Stephie! The new model works much better and closer models the DC conditions I get in reality over a spread of valves. Before I had to fudge the G2 circuit for DC conditions but now I don't Smile

I was attempting to model a DC coupled fully differential phono and have been messing around with an e-linear D3a/813 PP amp.

Thanks again!