Low frequency oscillation in Eiclone monoblock

a DIY, modified Mullard 5-20 monoblock design

Low frequency oscillation in Eiclone monoblock

Postby Greg Smith » Fri Mar 07, 2014 10:50 am

Hi,
One of my cathode-biased Eiclone monoblocks has taken to going into very low frequency oscillation lately. The oscillation usually starts about a minute after the amp is turned on. I noticed the woofer on that channel moving in and out, probably around 5hz. It isn't a huge cone excursion, but enough to be clearly noticeable - probably a few volts P-P. The amp still sounds normal, but clearly something is not right.

I built these amps point-to-point, not PCB. I need to crack them open and compare wiring layouts for obvious descrepancies. Otherwise I'm looking for likely causes of LF oscillation in an amp. Anyone?

The amps both worked normally for a few years. This is a recent development. I recently replaced the output tubes, but the oscillation occurs with both old and new tubes.

Cheers,
Greg
User avatar
Greg Smith
 
Posts: 30
Joined: Sun Dec 07, 2008 11:48 am
Location: Ottawa, Canada

Re: Low frequency oscillation in Eiclone monoblock

Postby Geek » Fri Mar 07, 2014 2:53 pm

Hi,

Given that it's just started, I'd say you have a bum power supply or decoupling cap.

Unless you have an ESR meter, change the lot of them with good quality caps from a reputable supplier.

Cheers!
-= Gregg =-
Fine wine comes in glass bottles, not plastic sacks. Therefore the finer electrons are also found in glass bottles.
User avatar
Geek
KT88
 
Posts: 3585
Joined: Sun Oct 21, 2007 3:01 am
Location: Chilliwack, British Columbia

Re: Low frequency oscillation in Eiclone monoblock

Postby Shannon Parks » Mon Mar 10, 2014 3:18 pm

Is the first stage DC coupled, or AC coupled (ie using a LM334Z in the LTP)?

Shannon
User avatar
Shannon Parks
Site Admin
 
Posts: 3764
Joined: Tue Mar 18, 2003 5:40 pm
Location: Poulsbo, Washington

Re: Low frequency oscillation in Eiclone monoblock

Postby Greg Smith » Mon Mar 17, 2014 1:45 pm

I'd be surprised if the coupling caps were leaking already, since I used brand new Orange Drops. The power supply filter caps were salvaged from a -48 volt SMPS unit for telecom switching equipment. They were high quality (Nichicon) units, but I suppose they could be faulty. Is it possible that the cathode bypass cap on the output tubes could also be the culprit?

I had been planning to convert these amps from cathode bias to fixed for some time. I'm going to go ahead with the conversion and examine the amps for bad caps at the same time.

Greg
User avatar
Greg Smith
 
Posts: 30
Joined: Sun Dec 07, 2008 11:48 am
Location: Ottawa, Canada

Re: Low frequency oscillation in Eiclone monoblock

Postby Greg Smith » Mon Mar 17, 2014 1:47 pm

Shannon, to answer your question, the first stage is DC coupled. Has that been identified as a problem in other builds?
Greg
User avatar
Greg Smith
 
Posts: 30
Joined: Sun Dec 07, 2008 11:48 am
Location: Ottawa, Canada

Re: Low frequency oscillation in Eiclone monoblock

Postby Shannon Parks » Tue Mar 18, 2014 12:54 pm

Greg Smith wrote:Shannon, to answer your question, the first stage is DC coupled. Has that been identified as a problem in other builds?
Greg


No, DC coupling improves low frequency stability so this is fine (conversely multiple AC-coupled stages with lots of feedback needs careful consideration).

Are you using the LM334Z or LM234Z? If so, what is the bias voltage across them?

Shannon
User avatar
Shannon Parks
Site Admin
 
Posts: 3764
Joined: Tue Mar 18, 2003 5:40 pm
Location: Poulsbo, Washington

Re: Low frequency oscillation in Eiclone monoblock

Postby Greg Smith » Tue Mar 18, 2014 12:59 pm

I am not using the LM334Z on the LTP - just a cathode resistor.
Greg
User avatar
Greg Smith
 
Posts: 30
Joined: Sun Dec 07, 2008 11:48 am
Location: Ottawa, Canada

Re: Low frequency oscillation in Eiclone monoblock

Postby Shannon Parks » Tue Mar 18, 2014 1:01 pm

Could you post some pix, Greg?

Shannon
User avatar
Shannon Parks
Site Admin
 
Posts: 3764
Joined: Tue Mar 18, 2003 5:40 pm
Location: Poulsbo, Washington

Re: Low frequency oscillation in Eiclone monoblock

Postby Greg Smith » Fri Dec 12, 2014 9:56 am

Nine months later, prompted by urgings from my wife (yes, you read that correctly), I've finally made time to investigate this problem.

I opened up the two monoblocks and compared them side-by-side to see if there was a discrepancy in the wiring/layout that would cause LF oscillation in one of them. Sure enough, the one that was oscillating is lacking a connection from the first filter capacitor ground (PT centre tap) to the chassis earth ground. Oddly enough, the amp functioned this way (with the amp's PSU/signal ground floating - the only earth ground would have been through the input patch cord shield to the preamp earth) for over two years before I noticed one of my woofer cones moving in and out, at about 5 Hz. I have not categorically determined that this is the fix but it seems like a likely cause. Any opinions?

In the meantime, I am in the process of converting both amps from cathode bias to fixed. (Why didn't I do this to begin with?) Shannon, I noticed that the Eiclone docs no longer include the cathode bias option schematics. Are you backing away from supporting this option? There is no real benefit to building with cathode bias - negligible difference in component cost, the cathode resistors generate a lot of heat, and you're wasting headroom).
User avatar
Greg Smith
 
Posts: 30
Joined: Sun Dec 07, 2008 11:48 am
Location: Ottawa, Canada

Re: Low frequency oscillation in Eiclone monoblock

Postby Shannon Parks » Sun Dec 14, 2014 7:35 am

Greg Smith wrote:In the meantime, I am in the process of converting both amps from cathode bias to fixed. (Why didn't I do this to begin with?) Shannon, I noticed that the Eiclone docs no longer include the cathode bias option schematics. Are you backing away from supporting this option? There is no real benefit to building with cathode bias - negligible difference in component cost, the cathode resistors generate a lot of heat, and you're wasting headroom).


When I updated the Eiclone manual last year I tried to prune it down and make it less messy (and also to not have to rev/update two BOMs and schematics), so I ditched the cathode bias part. I can email anyone the old manual if requested.

FWIW, the original pair of Eiclones are in my Uncle's workshop and they are indeed cathode biased. Maybe ten years old with daily operation during that time and over 10,000 operating hours.

Shannon
User avatar
Shannon Parks
Site Admin
 
Posts: 3764
Joined: Tue Mar 18, 2003 5:40 pm
Location: Poulsbo, Washington

Re: Low frequency oscillation in Eiclone monoblock

Postby Greg Smith » Tue Dec 30, 2014 12:32 pm

So I completed converting the amps to fixed bias, and fixed the grounding problem in one amp. No further problems with LF oscillation. I'm guessing the grounding problem was the cause. Set the bias to 38 mA per tube. Back in business!
User avatar
Greg Smith
 
Posts: 30
Joined: Sun Dec 07, 2008 11:48 am
Location: Ottawa, Canada


Return to eiclone

Who is online

Users browsing this forum: No registered users and 9 guests