OTL Asylum

OTL, Output Transformerless Amplifier User Group.

Return to OTL Asylum


Message Sort: Post Order or Asylum Reverse Threaded

Finally got my OTL apart to check suspected bias shift in one channel

164.54.56.3

Posted on October 28, 2004 at 09:00:55
Tom Dawson
Audiophile

Posts: 1882
Joined: May 14, 2000
And sure enough, that's what it was. That channel's output stage bias had dropped to -80V whereas the other one was still right at -62V, within 5% of where I set it at least five years ago. It appears likely that pushing the OTL hard into the 4-5 ohm load of the Basement Blasters may be precipitating parameter shift in one or more of the tubes in that channel (being fully dc coupled, everything potentially interacts here & many of the tubes have been operating in the amp for the full 16 years I've been running it), although the other channel's bias has remained unchanged.

 

Hide full thread outline!
    ...
Re: Finally got my OTL apart to check suspected bias shift in one channel, posted on October 28, 2004 at 10:07:53
Lew
Audiophile

Posts: 10911
Location: Bethesda, Maryland
Joined: December 11, 2000
I think that as tubes age, the grid to cathode voltage difference (i.e., the bias voltage) tends to go toward zero, if all other factors that determine bias remain constant. So your finding that the bias voltage has drifted "upward" (i.e., to a higher negative number) does not suggest to me that the output tubes are failing. Rather, I think something has changed upstream from the output tubes that are exhibiting a higher bias voltage.

 

Re: Finally got my OTL apart to check suspected bias shift in one channel, posted on October 28, 2004 at 14:08:46
Legendre
Audiophile

Posts: 503
Location: Saint Paul
Joined: October 22, 2004
If this is an Atma-Sphere style circuit, the output section bias is derived from the cathode voltage of the driver tube; 12AU7 in the old days, 6SN7 today.

Compare plate, grid and cathode voltages between the driver tubes in each channel - mainly the grid voltages. Take some notes, then swap drivers between channels and compare the results.

It's a rather simple circuit, so if voltages are way off (20V or so in your case) it's fairly simple to track it down; bad resistors, bad coupling caps, etc. And of course, the condition of the driver tube is critical to maintaining stable bias.

Bill

 

Re: Finally got my OTL apart to check suspected bias shift in one channel, posted on October 29, 2004 at 09:59:26
Tom Dawson
Audiophile

Posts: 1882
Joined: May 14, 2000
Thanks for responses. I tend to agree wrt the suggestion that the driver stage tube is the most likely culprit, and I plan to swap input & driver tubes between channels to isolate the problem before I rebuild the bias & offset adjustment circuits to make them externally adjustable this weekend. My OTL, btw, only has signal path capacitors for a couple of small value resistor bypasses in the DC level shifter and, in the feedback loops for phase margin compensation and to roll off to unity gain at DC.

 

Updates, posted on November 5, 2004 at 08:48:26
Tom Dawson
Audiophile

Posts: 1882
Joined: May 14, 2000
Well, I swapped all the input and driver tubes between channels & the bias for each barely changed. At this point, the suspect is the bias adjust pot itself (which I'll soon replace when I upgrade this circuit).

Update:

Last night, I finally got the bias and offset circuits properly rebuilt in the OTL and accessible from the back panel along with a bias monitor point for a hi-z voltmeter lead. For virtually the whole 16 year existence of this OTL, the circuit was a horrible, nearly inaccessible wad of parts including paralleled resistors, pots dangling from one lead, etc, which may be excusable for a prototype which the OTL was originally, but not for a long term use amp, and now that's all cleaned up and made more reliable. That's also one reason I didn't adjust those settings for the last five years, even when I swapped tubes:)

IAC, when I subsequently set the proper operating points, the OTL did a great job into the 3-5 ohm to about 500 hz Basement Blasters within its power envelope, with a midrange and high end miles beyond my other commercial amps being both more natural and more detailed with very noticeable depth imaging and occasionally being able to image beyond the speaker laterally, no mean achievement from DVDA & SACD in an acoustically untreated basement with stock JBL pro drivers used in the speakers. The bass was extremely good and fully integrated into the overall sonic presentation, also, but was subjectively leaner than that of the high damping factor Crown K1 & K2 I have while still having full extension. In fact, it probably is the extremely low bass phase shift the OTL has that contributes to its comparative bass character as much as anything, since the OTL has a first order shelving filter to DC unity gain and a 3db down point of 1hz on top of a LF damping factor in the hundreds, whereas the Crown amps have a third order rolloff with a corner frequency of 8hz.

Further update:

Listening to the OTL last night, I tried some impromptu tube mechanical damping that significantly cleaned up the sound of the OTL at higher SPLs (we're talking 110-115db peaks at the amp). Looks like I'll be looking at developing a permanent damping solution because the improvement is definitely worthwhile, giving better microdynamics at live SPLs, and the slight diffuseness I noticed earlier in the bass at higher SPLs also disappeared, it now hitting at just as hard (within the OTLs power envelope) as with the Crowns while being cleaner, tighter and better integrated with the rest of the soundstage than the Crowns (as previously noted). This makes sense, because it's below a few hundred hz where mechanical vibrations will most affect tubes such as 6as7gs.

 

Page processed in 0.020 seconds.