anyone want to help debug a weird problem?

General discussion area for tube amps.

Moderators: pompeiisneaks, Colossal

Post Reply
krash
Posts: 270
Joined: Mon Mar 20, 2006 12:43 am
Location: Cedar Park, TX
Contact:

anyone want to help debug a weird problem?

Post by krash »

OK,

I have a low-watt amp I have built, and on the output I have a switching jack. When you unplug the speaker, it switches to an internal reactive load which consists of a speaker motor (no cone) in series with an 8.2 ohm power resistor.

The amp circuit is sort of like a BF Princeton Reverb with no reverb, and of course with a 12AU7 power tube and all of the other little tricks I do to make a low-watt amp work, including gain adjustments etc. Only reason this is important to note is that it has an 820 ohm negative feedback coming off of the OT output, going into the cathode of the last gain stage before the cathodyne PI just like a Princeton Reverb (can't account for the values, I tuned this by ear).

Anyway, plugged into a speaker the amp is glorious. But when I unplug the speaker and it switches to the internal load, it motorboats and oscillates.

in parallel with the internal 16 ohm load is a 220K resistor, series with a 50K pot, wiper is going off to a line out jack.

I have jumpered the internal load onto the speaker jack and it does exactly the same thing. Motorboats.

Going to try it with an external passive load and see what happens. Just weird! Anyone have an idea?
-josh
--
Revelation Guitar Amplifiers
http://www.revelationamps.com
mbeldyga
Posts: 44
Joined: Sat Feb 06, 2010 4:55 pm

Re: anyone want to help debug a weird problem?

Post by mbeldyga »

I can't help you with your problem, but i think every trick that have to prevent damage OT made in switching jack (just like your method) isn't best solution. It's because when your cable from amp to cabinet or speaker coil will get damaged these methods just won't work because your jack is still in socket - fake load is not connected.
krash
Posts: 270
Joined: Mon Mar 20, 2006 12:43 am
Location: Cedar Park, TX
Contact:

Re: anyone want to help debug a weird problem?

Post by krash »

well, I fixed the problem by reversing the OT leads. Dunno why it didn't motorboat on the speaker but only the internal load, but regardless, it was NFB polarity causing the issue. This is the first amp I've ever built with negative feedback.

I am not doing this to protect the output, BTW. The amp is going to be used primarily plugged in direct using a cabinet modeler, so the internal load is there to provide a line out. I doubt the guy buying the amp will ever plug it into an external speaker. In reality the external speaker jack is there for a convenience, not the other way around.
-josh
--
Revelation Guitar Amplifiers
http://www.revelationamps.com
User avatar
Bob-I
Posts: 3791
Joined: Thu Feb 02, 2006 10:06 pm
Location: Hillsborough NJ

Re: anyone want to help debug a weird problem?

Post by Bob-I »

krash wrote:well, I fixed the problem by reversing the OT leads. Dunno why it didn't motorboat on the speaker but only the internal load, but regardless, it was NFB polarity causing the issue. This is the first amp I've ever built with negative feedback.
Sounds to me like the NFB isn't connecting when you're using the speaker out. I'd double check the wiring of the NFB and output jacks.
Rick
Posts: 151
Joined: Sat Mar 24, 2007 11:46 pm

Re: anyone want to help debug a weird problem?

Post by Rick »

Yes, when the plate to OT connections are not correct, may PP amps will oscillate and any NFB will actually be positive feedback exacerbating the symptoms. If disconnecting the alleged NFB wire improves things, that is a telltale that plate leads are reversed. It can occur in open loop amps or amps with NFB. Reversing the connections to long tail phase inverter coupling will have the same effect. With cathodyne, I am not so familiar, but anyway, when howling and noise are excessive inexplicably, it's worth switching the OT plate leads to see if that is causing it.

Using unknown trannys gives you a 50/50 chance so it's pretty common, sometimes going by the wiring color will still not be correct and they need switching (according to layouts), perhaps they are connected opposite or the builder doesn't follow the color coding scheme. I've had some open loop amps that actually performed fairly well but just weren't quite right and it turned out to be the OT plates were reversed, the amp improved noticeably but it was strange it worked so well reversed that it wasn't immediately obvious this was the case.
krash
Posts: 270
Joined: Mon Mar 20, 2006 12:43 am
Location: Cedar Park, TX
Contact:

Re: anyone want to help debug a weird problem?

Post by krash »

well ya know, I never build amps with negative feedback. customer special. anyhoo, this is the first amp I built with this particular OT, so I am sure I just got the wiring color code wrong. sounds good now.

later-
-josh
--
Revelation Guitar Amplifiers
http://www.revelationamps.com
mlp-mx6
Posts: 1111
Joined: Sat Feb 04, 2006 4:24 pm
Location: NW Atlanta

Re: anyone want to help debug a weird problem?

Post by mlp-mx6 »

I always clip the OT/plate wires in place at first to make sure I have to polarity right. Once I'm sure, only THEN will I cut them to length and solder.
Wife: How many amps do you need?
Me: Just one more...
User avatar
FUCHSAUDIO
Posts: 1256
Joined: Fri Apr 08, 2005 2:48 pm
Location: New Jersey (you got a problem with that ?)
Contact:

Re: anyone want to help debug a weird problem?

Post by FUCHSAUDIO »

When the amp is working, connect then disconnect the nfb. If it's wired right the gain goes DOWN when the lead is connected. That determines phase is correct..
Proud holder of US Patent # 7336165.
Rick
Posts: 151
Joined: Sat Mar 24, 2007 11:46 pm

Re: anyone want to help debug a weird problem?

Post by Rick »

FUCHSAUDIO wrote:When the amp is working, connect then disconnect the nfb. If it's wired right the gain goes DOWN when the lead is connected. That determines phase is correct..
So will the howling, squealing, and hiss if that is present.
Post Reply