Bifilar IT with signal on both ends of secondary

Design and use of the various types.
Post Reply
reVintage
Posts: 365
Joined: Fri Sep 07, 2007 9:54 am
Location: Sweden

Bifilar IT with signal on both ends of secondary

Post by reVintage »

Hey Dave,

Now we´ve come to the right place! Tried the link in my computer and it worked OK. Hope it works better for you here.

This is close the textbook Schade. Personally I would choose another driver but it´s not important here.
Attachments
Schade6L65687IT.asc
(4.22 KiB) Downloaded 509 times
Schade.PNG
Schade.PNG (42.32 KiB) Viewed 14037 times
Brgds
Lars
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

I'm on a mac so i'm sure that had something to do with it.

I just got a netlist in a text window. Tried saving it as an .asc and it didn't work. I don't want to hijack the Asylum thread so I'll post there and when it falls off the page we can continue here.

dave
reVintage
Posts: 365
Joined: Fri Sep 07, 2007 9:54 am
Location: Sweden

Post by reVintage »

OK, just a quickie before going back to AA: What capacitance can we estimate in a IT like this?
Brgds
Lars
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

I just measured 20n on a big bifilar that I have here. The results were not pretty.
reVintage
Posts: 365
Joined: Fri Sep 07, 2007 9:54 am
Location: Sweden

Post by reVintage »

Na, this was a deadend. Hope the guy hasn´t a bifilar.
Brgds
Lars
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

I'm not convinced the bifilar is a problem, but looking at the overload characteristics and the idea that you need 70V peak from the 5687 to deliver 10V peak to the grid of the 807 makes me agree with your assessment.

dave
reVintage
Posts: 365
Joined: Fri Sep 07, 2007 9:54 am
Location: Sweden

Post by reVintage »

Hey Dave,

Yes, I now you´re not fond of pentodes! This one sims like a dream (hmmm almost). There is a BIG problem though! Major resonance in the low end! Play around and tell me what you think.
Attachments
Schade6L6D3AIT.asc
(4.28 KiB) Downloaded 441 times
Brgds
Lars
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

Interesting. I didn't run it yet but figured a picture might interest the rest.
Attachments
Screen shot 2009-11-13 at 11.48.57 AM.png
Screen shot 2009-11-13 at 11.48.57 AM.png (53.66 KiB) Viewed 13979 times
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

Damn you lars... I have work to do stop making me think!

I kept staring at the load on the D3A and seeing the 100hy in parallel with the feedback. By instability at LF are you referring to the 5-6dB peak at 10hz? It looks like a typical LC resonance but there aren't any C's (coupling or bypass) to cause this.

I keep looking at the load to the D3a and we know a finite inductive load will not work so there needs to be a parallel load for the 100hy's to "work against". The load appears to be the 68K in series with Rp || Rl of the 807 or about 68K which again is nowhere near low enough for the 100hy's to work against. Of course the feedback resistor has current through it and this must be an "active load" that is in parallel with the 100HY to load the tube at low frequency. when you probe the current through it you get like 20ma at 10hz. 20ma through 68K is 1360V so i'm scratching my head. (switching to linear mode for real units helps) Since I highly doubt that those numbers will ever happen in the real world, I wondered what happens if we put a resistor across the primary of the IT to better define the load on the d3a. 10K goes a long way to making things seem to behave.

BTW, are there any other good pentode models out there besides the D3A? I'd love to find a 7788 and a 7722.

dave
reVintage
Posts: 365
Joined: Fri Sep 07, 2007 9:54 am
Location: Sweden

Post by reVintage »

Hey Dave,
The load is 2-4kohm due to the plate-grid feedback loop around the 807. So the D3a is properly loaded, but not by the transformer. But I have to admit I am not fully aware of how this one works. Got the idea from Michael Koster using a MOSFET as complement for driving A2.

About Spicemodels they seem rare. Should really like to find E280F and
C3m.

Is this one acceptable? The peak is dependant on Lprimary of the OPT but can be damped somewhat by fiddling with the screencapacitor of the driver.
Attachments
Schade6L6D3AIT.asc
(4.2 KiB) Downloaded 460 times
Brgds
Lars
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

Look what happens when you change the reflected load. Granted most speakers are well beyond the bass impedance peak at 10hz but I try to stay away from circuits that need to be tuned, particularly when using nonlinear devices.

I still like the first sketch with a 10K across the primary of the IT. It still has the reflected load issues but the response curves look "less complex" I still wonder about stability though.

dave
lanshan75
Posts: 6
Joined: Wed May 16, 2012 3:37 am

Post by lanshan75 »

Hello,everybody,I am just a new member of this forum,I want to make more friends here,I hope you can like me.
Post Reply