Strange hum in a Williamson amp

vlucian · 24th November 2010, 09:57 AM

Hi there.
I have the following situation:
- an williamson amplifier with 6p3s final tubes
- initially the amp was built with 2 small output transformers without ultraliniar windings but latter I decided to replace them with bigger ones (also having UL)
- I had to change the box (the new transformers are 4 times bigger) and to rewire everything
- in the end everything was fine excepting for a low hum not very disturbing but still audible
- I tried few things including changing the wires or movement of the power transformer outside the box only to see if a wrong positioning was the cause: the hum was still there no matter what
- I must mention I used a star ground pattern
- I went to a friend who has an oscilloscope and we made some measurements: we found a very high frequency (around 1 megahertz)!! The oscilloscope was an old Tesla one so we didn't had to much measurement possibilities (like harmonics).

What do you think should I try? Maybe there are not much info, please ask if something is not clear.
thanks a lot!

DF96 · 24th November 2010, 03:26 PM

Oscillation can sometimes happen with UL. If not already present, try adding grid stoppers to the output control grids. Then g2 stoppers (say, 470ohms or 1K 2W wirewound). Finally, you may have to add snubber networks from each anode to g2.

Svein_B · 24th November 2010, 04:47 PM

Try disconnecting the global feedback until you are sure everything is OK without it.
If stable without GNFB you may need to tweak the feedback loop, or some other FR compensation to get it stable with feedback.

The Williamson can sometimes be tricky to get right.

Svein

DF96 · 24th November 2010, 05:01 PM

1MHz oscillation is unlikely to be due to the global feedback loop. More likely to be local feedback in the output stage due to parasitics in the OPT.

However, disconnecting the global loop is a good debugging technique as it stops interactions complicating things.

roline · 25th November 2010, 10:28 PM

I'd look for possible ground loops but star ground really should prevent it. I had to increase my power supply cap ( double the size) to rid me of a very low level hum to a totally inaudible level.
If it is a circuit instability, try to jumper 1,000 pf to the input pins of the tubes and see if the hum still exists. Move to the next input etc... This should force a dominant pole into the k-hz range and stop high frequency oscillation. If it works, you know the area of the circuit to investigate. (without the global feed back) Just thinking, could you have connected the output feed back inverted?. Does it get louder of quieter with the global feedback?
In the attached schematic R7, C5 create the dominate pole for Amp's stability. The capacitor can be increased for stability but excessively large with decrease the power bandwidth.

vlucian · 27th November 2010, 07:13 PM

Thank you all for your ideas and suggestions. Next week I'll go to my friend again (I let the amp there) to try some things especially with the feedback. I'll let you know the results.
thanks again

vlucian · 2nd December 2010, 09:03 PM

So, we did the following:
- changed the resistor in the NFB from 10k to 20K => the hum was louder
- changed the resistor with a potentiometer and try to obtain minimum audible hum => the new value is 2k5
- view on the scope: the now hardly audible hum still has high frequency component with about 200mV amplitude; rotating potentiometer changed the noise level but also cahnged the frequency
-disconnecting the global feedback only bring more noise
- we didn't changed the feedback capacitor value
I wonder there is no way to cut completly that noise. What else sgould I do?
thanks

roline · 3rd December 2010, 10:20 AM

Disconnect the coupling caps from the driver stage and jumper them to ground. Does the output still have the instability? If it does, then it is a combination of the bias and output tubes. Also disconnect the GNFB and look at the output of the driver section.
With the input grounded, it should be quiet.
The GNFB was trying to overcome it, just need to find the source.

DF96 · 3rd December 2010, 12:37 PM

The feedback resistor reduces the hum simply by reducing the gain. It doesn't solve the problem, just hides it.

Can you sketch an exact diagram of your star grounding? It could be charging pulses getting into the audio. This only needs one wire connected to the wrong place, which could be very close to the right place!

kevinkr · 3rd December 2010, 06:49 PM

Fix your RF oscillation problem first, the hum problem may then go away. (From hard experience)

24th November 2010, 09:57 AM	#1
vlucian diyAudio Member Join Date: Mar 2005 Location: bucharest	Strange hum in a Williamson amp Hi there. I have the following situation: - an williamson amplifier with 6p3s final tubes - initially the amp was built with 2 small output transformers without ultraliniar windings but latter I decided to replace them with bigger ones (also having UL) - I had to change the box (the new transformers are 4 times bigger) and to rewire everything - in the end everything was fine excepting for a low hum not very disturbing but still audible - I tried few things including changing the wires or movement of the power transformer outside the box only to see if a wrong positioning was the cause: the hum was still there no matter what - I must mention I used a star ground pattern - I went to a friend who has an oscilloscope and we made some measurements: we found a very high frequency (around 1 megahertz)!! The oscilloscope was an old Tesla one so we didn't had to much measurement possibilities (like harmonics). What do you think should I try? Maybe there are not much info, please ask if something is not clear. thanks a lot!

3rd December 2010, 06:49 PM	#10
kevinkr diyAudio Moderator Join Date: Sep 2004 Location: Boston, Massachusetts Blog Entries: 6	Fix your RF oscillation problem first, the hum problem may then go away. (From hard experience) __________________ "Extraordinary claims require extraordinary evidence." - Carl Sagan

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
My KT88 Williamson Amp Build	Loren42	Tubes / Valves	64	11th March 2013 03:53 PM
Strange transformer hum	Atilla	Power Supplies	12	1st January 2010 01:12 AM
Strange circuit - heater hum opt feedback?	m6tt	Tubes / Valves	3	23rd April 2008 06:13 PM
Strange Hum	pftrvlr	Tubes / Valves	4	20th February 2008 12:02 PM
strange hum problem	zeus_threat	Solid State	8	18th December 2007 02:55 PM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read


	Please consider donating to help us continue to serve you. Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving

24th November 2010, 03:26 PM	#2
DF96 diyAudio Member Join Date: May 2007	Oscillation can sometimes happen with UL. If not already present, try adding grid stoppers to the output control grids. Then g2 stoppers (say, 470ohms or 1K 2W wirewound). Finally, you may have to add snubber networks from each anode to g2.

24th November 2010, 04:47 PM	#3
Svein_B diyAudio Member Join Date: Oct 2005	Try disconnecting the global feedback until you are sure everything is OK without it. If stable without GNFB you may need to tweak the feedback loop, or some other FR compensation to get it stable with feedback. The Williamson can sometimes be tricky to get right. Svein

24th November 2010, 05:01 PM	#4
DF96 diyAudio Member Join Date: May 2007	1MHz oscillation is unlikely to be due to the global feedback loop. More likely to be local feedback in the output stage due to parasitics in the OPT. However, disconnecting the global loop is a good debugging technique as it stops interactions complicating things.

27th November 2010, 07:13 PM	#6
vlucian diyAudio Member Join Date: Mar 2005 Location: bucharest	Thank you all for your ideas and suggestions. Next week I'll go to my friend again (I let the amp there) to try some things especially with the feedback. I'll let you know the results. thanks again

2nd December 2010, 09:03 PM	#7
vlucian diyAudio Member Join Date: Mar 2005 Location: bucharest	So, we did the following: - changed the resistor in the NFB from 10k to 20K => the hum was louder - changed the resistor with a potentiometer and try to obtain minimum audible hum => the new value is 2k5 - view on the scope: the now hardly audible hum still has high frequency component with about 200mV amplitude; rotating potentiometer changed the noise level but also cahnged the frequency -disconnecting the global feedback only bring more noise - we didn't changed the feedback capacitor value I wonder there is no way to cut completly that noise. What else sgould I do? thanks

3rd December 2010, 10:20 AM	#8
roline diyAudio Member Join Date: Sep 2010 Location: East Tennessee	Disconnect the coupling caps from the driver stage and jumper them to ground. Does the output still have the instability? If it does, then it is a combination of the bias and output tubes. Also disconnect the GNFB and look at the output of the driver section. With the input grounded, it should be quiet. The GNFB was trying to overcome it, just need to find the source.

3rd December 2010, 12:37 PM	#9
DF96 diyAudio Member Join Date: May 2007	The feedback resistor reduces the hum simply by reducing the gain. It doesn't solve the problem, just hides it. Can you sketch an exact diagram of your star grounding? It could be charging pulses getting into the audio. This only needs one wire connected to the wrong place, which could be very close to the right place!