Motor boating Hoffman half-AC-30

[dbroekema]

Hi guys,

This amp is built on a Hoffman AC30 board, but only has the TB channel and 2xEL84 output section. Using the Heyboer 18 watt tranny set from Doug's page and an 100 Ma x 10Hn Hammond choke. Amp is in a 5e3 type chassis and cab with a single Weber 12".

Anyhow, the thing sounds awesome, except if I have the bass control set too high it sometimes goes into a motor boating LF oscillation, especially when overdriving it with a pedal. This can be shut down by turing the bass control full-off and then resetting. I've looked over the wiring, chopsticked it, checked grounds- no luck. There's about 320V on the plates of the EL84's with a 5V4 recto and around 300V when I run a 5R4GB.

Any ideas on what i should look for? Also any mods that might change this? Thinking that changing the .1uf caps in the phase inverter to .022uf might knock off some bass, could that help?
Thanks!
Dave

[sluckey]

What happens if you put a 220K resistor between the junction of R20/C9 and ground?

Did you use new filter caps? Try bridging another filter cap across each of your existing filter caps while the amp is motor-boating.

[dbroekema]

Slucky wrote:
"What happens if you put a 220K resistor between the junction of R20/C9 and ground?"
-Should I have this? I don't have the "normal" channel, so I don't have the another 220k connected to ground (through the normal channel's volume pot). But this is not the first amp like this I've built, with the single pre-amp channel, and in the past I've even omitted the 220K altogether without troubles. Does the circuit require this leak to ground entering the PI?

"Did you use new filter caps? Try bridging another filter cap across each of your existing filter caps while the amp is motor-boating."
-I did use new filter caps, generally higher than the stock ratings (I believe 16/8ufs, I've got 20 of 40 uf in there), and I have an additional dropping resistor (250ohm) and cap (33uf) node directly off the rectifier before it sees the B+ rail. The purpose of this was to get voltage a little bit "safer" for the power tubes, but filtering should be better than normal.

Curious on your thoughts about your first suggestion. Thanks for the reply!

[sluckey]

Does the circuit require this leak to ground entering the PI?

Leak to ground is not important. But the 2:1 voltage divider that would exist if you had a normal channel with it's volume on zero may be. I think it's worth a try. Easy to do/undo. Doesn't have to be exactly 220K.

[octal]

I'm wondering about the extra 250 ohms and 33uf cap. Seems like adding that much sag could contribute to motorboating. (Ie output tubes draw current, B+ is pulled down, the other power supply nodes down the line are also pulled down and depending on the time constants of the various nodes, it seems you might wind up with a positive feedback condition.)

[dbroekema]

Octal wrote:
"I'm wondering about the extra 250 ohms and 33uf cap. Seems like adding that much sag could contribute to motorboating. (Ie output tubes draw current, B+ is pulled down, the other power supply nodes down the line are also pulled down and depending on the time constants of the various nodes, it seems you might wind up with a positive feedback condition.)"

-Actually the motor boating was there from the start, before I put this extra node in. Never experienced this before in any build so really interested in any replies, thanks.

[dbroekema]

Slucky wrote:
"Leak to ground is not important. But the 2:1 voltage divider that would exist if you had a normal channel with it's volume on zero may be. I think it's worth a try. Easy to do/undo. Doesn't have to be exactly 220K."

-Awesome, I'll try this. Out of curiosity, how high could I go with the value on this resistor? I want to cure the motor boating, but I also don't want to lose signal here of at all possible. 220k, 470K, 1M?
Thanks!

[HotBluePlates]

... Out of curiosity, how high could I go with the value on this resistor? I want to cure the motor boating, but I also don't want to lose signal here of at all possible. ...

I'd suggest just starting with the 220kΩ which is close to mimicking the stock circuit condition. Sluckey is suggesting adding it just to see if it helps; if it does, you can optimize later.

[dbroekema]

Cool, thanks guys, I'll try this and post back as soon as I can sneak away from the family long enough to do some soldering!

[dbroekema]

That did indeed fix the motor boating, thanks Slucky. However, it does sound (at first, very quick listen) that I may have lost some gain and some bite in the top end. But that seems strange, since as you mentioned, the set-up now is the same as a "complete" amp with normal channel volume on zero. I might experiment with that resistor value and see what happens.

Thanks again!

[dbroekema]

Still stable with a 1M in there, so that's how it stays. Thanks again, all.