Non linear power

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi All, I recently acquired a vintage BGW 100B power amplifier that, for all intents and purposes works fine but I find that the 2 channels do not follow each other gain wise until both pots are full up, anywhere below that say at 9,12 or 3 o'clock positions the output is unequal.
The gain pots are identical, scratch and dropout free through out their travel however, the output transistors on one channel have been replaced at some point.
The OEM outputs are a Motorola SJ9215 and SJ9216 pair and the serviced channel is now equipped with a Toshiba 2SB 554 and 2SC 424 pair.

Can that be the cause of the difference or is there something more sinister at play?
 
Administrator
Joined 2007
Paid Member
If the gain is equal at full volume then the most likely suspects are in fact the pots. Even dual gang pots often track badly.

Out of all the components in an amplifier, it is the feedback network values that determine gain (so just perhaps four parts out of the whole thing). The transistors have no effect on overall gain.

The easiest check on the pots is to apply a mono test tone (sine) and look with a scope at the output on the wiper of each.

Oddball possibilities could be a changed or incorrectly replaced part right after the pot such that the input impedance of the amp had been effectively altered.
 
I`ll give the wipers a measure. It`s worth mentioning that this amp has seen a fair amount of tar and nicotine in it`s lifetime, I have yet to Deoxit so perhaps that may solve it.
Like I said, the amp does perform normally otherwise and both sides are within millivolts of each other at waveform clipping.
I just find this type of condition odd, I`ve never seen it before actually.
 

Attachments

  • BGW 100B Manual.pdf
    807.9 KB · Views: 38
Administrator
Joined 2007
Paid Member
Yes, see what the wiper voltages are (I take it you have a scope ?). If not then you can still do it using a decent DVM as long as it can measure low AC voltage. Use a test tone of around 400Hz in that case as that is within the frequency range of any meter.

Have a look around the mono switching as well. If the amp is really gungy then anything could be happening with regard to switches.
 
UPDATE

I had an opportunity to do some work on the amp the other day, started with spraying down the PCB with pure alcohol, agitating and rinsing with demineralised water, cleaned (D5) the pots and switches, all the input jacks, octal sockets and jumper plugs as well as the pin and sleeve type connectors that populate the motherboard i.e. B+/B-, grounds etc. The power cable needed some attention too as the jacket was becoming frayed right at the strain relief and the chassis was drilled and tapped to accommodate new hardware for the top and bottom covers.
After reassembling and passing the DBT came the listening test. To my surprise, both left and right channels were equal and tracked in perfect harmony.....for the first couple of watts and then the left channel went into clipping . No fuses were blowing and nothing is heating up so I figured it cant be catastrophic. The scope reveals positive cycle clipping at only 5vPP, damn. It didn’t take long to find a cold solder joint on I believe was the base of the replacement NPN Toshiba output. The emitter looked equally questionable as did the PNP counterpart so they were all re flowed and just like that, problem solved. I`m guessing that it was only the driver section that was running the speaker up until the premature clipping point, in any case I was just happy that everything is working correctly and sounding very good.
Power checked out at 64.53wpc@8, 79.35wpc@4 and 82.67wpc@3, 1khz both channels driven just below clipping. Nice little amp :)
 

Attachments

  • 14.jpg
    14.jpg
    771 KB · Views: 66
Administrator
Joined 2007
Paid Member
Excellent, well done :up: It sounds a bit like there could have been two issues at work here and that the cleaning of it all has fixed one, and then this other problem appeared.

Its quite satisfying to resurrect stuff like this and give it all a good clean.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.