DC coupling nightmare

All the parts have to be put together at some point. Have fun!
Post Reply
Daniel
Posts: 45
Joined: Fri Jan 10, 2014 4:50 pm

DC coupling nightmare

Post by Daniel »

Hi everyone,
I tried to get a dc amp to work with stacked supplies. After many failures and cap coupling to listen to music I played around again, had a breakthrough and more failures.
I am a muppet when it comes to Spice and all things computers, I just used it to draw circuits.
I started with this combination of signal circuit and PSU which produced earsplitting noise:
Image
Image

I then disconnected the 530V supply, put in a floating external PSU and changed the signal part to this:
Image

This hummed more than acceptable but I could listen to music and it sounded WAY better than the RC coupled version. So I went on and cleaned up the anode supply to this, sticking to the external supply for the 801:
Image

With this I am back to earsplitting noise where I started. Do any of you have an idea what is going on here?

Thanks for any help,
Daniel
Attachments
211 stacked original.asc
(2.6 KiB) Downloaded 212 times
211 PSU cap couple.asc
(4.28 KiB) Downloaded 224 times
211 stacked external supply.asc
(2.63 KiB) Downloaded 207 times
211 PSU dc cleaned up.asc
(3.63 KiB) Downloaded 193 times
Daniel
Posts: 45
Joined: Fri Jan 10, 2014 4:50 pm

Re: DC coupling nightmare

Post by Daniel »

I put in a 470uF capacitor to bypass V1. Still noisy as hell.
dave slagle
Posts: 2092
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

Hey.

I loaded it up and had to make a few changes to get it working.

I am guessing that L2/C1 L3/C5 and L4/C2 represent the last filter section of the PS???

dave
Attachments
Screen shot 2015-05-29 at 2.53.34 PM.png
Screen shot 2015-05-29 at 2.53.34 PM.png (18.98 KiB) Viewed 6494 times
Get Your Fix
www.hifiheroin.com
Daniel
Posts: 45
Joined: Fri Jan 10, 2014 4:50 pm

Post by Daniel »

dave slagle wrote:Hey.

I loaded it up and had to make a few changes to get it working.

I am guessing that L2/C1 L3/C5 and L4/C2 represent the last filter section of the PS???

dave
Hi Dave,
thanks for checking in. Filament supplies (DC, LCL filtered) are external and so is the anode supply. L2/C1 L3/C5 and L4/C2 are the last filter section placed in the signal part.

Could you do me a favor and upload the .asc file you used?
dave slagle
Posts: 2092
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

.asc is attached. I use Stephies drop down models so if you want to use an .inc file you will need to delete the triodes and replace them with the appropriate symbol.

dave
Attachments
211 stacked.asc
(4.43 KiB) Downloaded 215 times
Get Your Fix
www.hifiheroin.com
Daniel
Posts: 45
Joined: Fri Jan 10, 2014 4:50 pm

Post by Daniel »

Thanks a bunch! One more thing to turn this into Spice kindergarten: I simulated many versions of that circuit and got more or less these two different waveforms:
Image
Image

Is the second waveform telling me that this circuit will give me the noise I encountered?
Daniel
Posts: 45
Joined: Fri Jan 10, 2014 4:50 pm

Post by Daniel »

OK, I found the node measurement and also a way to switch this mess on and off in the simulation:
Image
dave slagle
Posts: 2092
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

Hey,

I'm not sure what the pictures are showing me. Can you post the capture with both the schematic and the frequency response graph?

also not sure why your first two pictures didn't show up. You can simply add the image file to the post then it will be hosted by the forum software and always show up.

dave
Get Your Fix
www.hifiheroin.com
Daniel
Posts: 45
Joined: Fri Jan 10, 2014 4:50 pm

Post by Daniel »

I didn't find the screenshot of the circuit that gave the static but I'll recreate it later.
As the circuits that looked good in Spice still gave me more or less (but always way too much) static noise I was wondering if there are some rules of thumb where to connect the two power supplies.
It puzzles me that with this signal circuit
Image
that PSU gives me really loud static:

Image

but that other sounded really great with noise that came through in quiet passages and when no music was playing:
Image
dave slagle
Posts: 2092
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

hey,

i'm unclear which supply fared better. What is the reasoning behind the hookup pictured below. It is set up as a typical bipolar design (+- the same voltage) if the CT is "ground" but the difference in filters might cause issues. what part is grounded?

If it is intended to give a B+ and 1/2B+ then the bottom is grounded and the filters for each leg connect directly to ground.

dave
Attachments
Screen shot 2015-06-10 at 10.35.30 AM.png
Screen shot 2015-06-10 at 10.35.30 AM.png (32.17 KiB) Viewed 6468 times
Get Your Fix
www.hifiheroin.com
Daniel
Posts: 45
Joined: Fri Jan 10, 2014 4:50 pm

Post by Daniel »

Howdy!
You are coming back to the original question here. This was the original anode supply for the 211 and the 801A giving B+ and 1/2 B+. I never got this to work dc-coupling the Ba to the 801A (full output static noise), so I disconnected the 1/2 B+ to the 801A, put in a lab supply and the noise went down enough to enjoy the music.
Then I "cleaned up" the CT supply by disconnecting the center tap and taking out the 801 filter caps and the chokes in the ground leg. And was back to full on noise.
dave slagle wrote:hey,

i'm unclear which supply fared better. What is the reasoning behind the hookup pictured below. It is set up as a typical bipolar design (+- the same voltage) if the CT is "ground" but the difference in filters might cause issues. what part is grounded?

If it is intended to give a B+ and 1/2B+ then the bottom is grounded and the filters for each leg connect directly to ground.

dave
dave slagle
Posts: 2092
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

I think the shared ground of the 1/2 B+ full B+ arrangement prevents use as a stacked supply. You could do it as pictured but this will give you two nearly identical stacked voltages.


dave
Attachments
Screen shot 2015-06-10 at 2.20.34 PM.png
Screen shot 2015-06-10 at 2.20.34 PM.png (32.64 KiB) Viewed 6464 times
Get Your Fix
www.hifiheroin.com
Daniel
Posts: 45
Joined: Fri Jan 10, 2014 4:50 pm

Post by Daniel »

As I tried to explain earlier I am not using the 1/2 B+ arrangement anymore but put in another independent power supply.
Having the chokes in the center tap leg didn't work; for some reason this gave me a positive instead of a negative grid voltage on the 801.

When you build a dc amp with stacked supplies, how do you connect the two supplies? In the signal section or the last PSU cap or do you connect all the PSU caps?
dave slagle wrote:I think the shared ground of the 1/2 B+ full B+ arrangement prevents use as a stacked supply. You could do it as pictured but this will give you two nearly identical stacked voltages.


dave
Daniel
Posts: 45
Joined: Fri Jan 10, 2014 4:50 pm

Post by Daniel »

Putting in cathode resistor bypass capacitors brought some changes. On the Ba and 801A it did nothing, when installed on the 211's Rk the wideband noise turned into some motorboating at 2-3 Hz. Removing the original filament supply and hooking up a lab supply to each filament changed nothing.

I'll stay on it...
Post Reply