I've just finished re-capping a Harmony H-305. Here is the schematic and parts list.
Schematic
Parts list
The amp played fine for about 5 mins during testing, then all of a sudden the signal dropped out and the speaker started to emit a loud buzz. I noticed the plates of the rectifier (5Y3) starting to glow. I tested all the tubes - no shorts and good emission values. I swapped the 5Y3 for a NOS, same problem. The buzz/hum even happens with the pre and power tubes removed. With the rectifier removed I get normal values on the PT (255V per tap, 5.2V on the heater). With the rectifier I get ~ 4.7vac on the heaters and about 120vdc on the B+, but the plates start over conduct immediately.
Any ideas?
thanks!
Schematic
Parts list
The amp played fine for about 5 mins during testing, then all of a sudden the signal dropped out and the speaker started to emit a loud buzz. I noticed the plates of the rectifier (5Y3) starting to glow. I tested all the tubes - no shorts and good emission values. I swapped the 5Y3 for a NOS, same problem. The buzz/hum even happens with the pre and power tubes removed. With the rectifier removed I get normal values on the PT (255V per tap, 5.2V on the heater). With the rectifier I get ~ 4.7vac on the heaters and about 120vdc on the B+, but the plates start over conduct immediately.
Any ideas?
thanks!
Yeah, any chance one of your filter caps is wired backwards?
COme to think of it, with the power tubes removed, a bad filter would still make no sound. If you get sound from the speaker with NO power tubes, then something is dragging current through the output transformer, most likely a short from primary to frame or from primary to secondary
disconnect your output transformer B+ lead. Does it stop?
COme to think of it, with the power tubes removed, a bad filter would still make no sound. If you get sound from the speaker with NO power tubes, then something is dragging current through the output transformer, most likely a short from primary to frame or from primary to secondary
disconnect your output transformer B+ lead. Does it stop?
OK, so the cap is probably fine, unfortunately I suspect your output transformer is not.
Power off, power tubes removed, CT still disconnected. Now measure DC resistance to chassis from that center tap wire, and also from either plate wire. Are any of those three primary winding wires showing resistance to ground? If one end were shorted to frame, then the resistance of the winding wire would be put from B+ to ground. And check for resistance to the secondary too. A good transformer has no continuity from windings to frames or between windings.
Power off, power tubes removed, CT still disconnected. Now measure DC resistance to chassis from that center tap wire, and also from either plate wire. Are any of those three primary winding wires showing resistance to ground? If one end were shorted to frame, then the resistance of the winding wire would be put from B+ to ground. And check for resistance to the secondary too. A good transformer has no continuity from windings to frames or between windings.
CT to chassis: 825 ohms.
Primaries to chassis; 1K and 660 ohms.
With the negative lead of the meter to chassis, touching the pos meter probe to either the CT or 1K primary causes some barely audible static to be emitted by the speaker. Again, no tubes installed. Secondaries show infinite to ground. I should mention that when the amp did work, it was quite noisy unless the tone knob and both volumes were completely turned down. Even though the inputs are grounded (I cleaned the contacts and they show continuity) you'd get noise as soon as you turned up the volumes with nothing plugged in.
Primaries to chassis; 1K and 660 ohms.
With the negative lead of the meter to chassis, touching the pos meter probe to either the CT or 1K primary causes some barely audible static to be emitted by the speaker. Again, no tubes installed. Secondaries show infinite to ground. I should mention that when the amp did work, it was quite noisy unless the tone knob and both volumes were completely turned down. Even though the inputs are grounded (I cleaned the contacts and they show continuity) you'd get noise as soon as you turned up the volumes with nothing plugged in.
Circuit noise is a spearate issue from leaky OT. I am afraid your transformer is shot.
Your meter resistance scale runs a small current through a circuit to read the resistance, that small current causes the transformer to induce a little voltage in the secondary, which you hear out the speaker.
Your meter resistance scale runs a small current through a circuit to read the resistance, that small current causes the transformer to induce a little voltage in the secondary, which you hear out the speaker.
Before condemning the OPT, unsolder the wires that connect the OPT to the 6V6 sockets, and recheck the resistance readings. I have seen sockets fail due to moisture absorbtion. This type of failure usually results in smoke, stink, fireworks, or all of the above though.
Probably not, although the B+ voltage probably went up a bit with a new cap helping the failure along.
Probably not, although the B+ voltage probably went up a bit with a new cap helping the failure along.
Sure, we could have arcing sockets or other leakages, worth checking, unsolder two wires and find out. But I have to think the 600 ohms to frame wasn;t the sockets.
I agree the cap is probably not the cause. If you had played it a while longer with crummy caps it probably would have gone ahead and failed too.
I agree the cap is probably not the cause. If you had played it a while longer with crummy caps it probably would have gone ahead and failed too.
The older black phenolic sockets were prone to arcing which led to carbonization. This usually happened with dirty sockets that were stored in a moist environment. 600 ohms is a very real resistance reading for a carbonized socket, but there is uaually a tell tale smell. A socket could have arced years ago due to being operated without a speaker and short out later in life. I have seen this happen, but it is usually in higher powered amps using 6L6GC tubes.
The arc starts from pin 3 (plate) to pin 2 (heater) which is often grounded. Look carefully with a magnifier for a burned track. It will be black on black which is hard to find. The burnt stuff can be scraped away and the socket cleaned with WD40 and a toothbrush, but it may fail again.
I agree the OPT is probably the culprit, and moisture related arcing is probably the cause. Still, the tube sochets should be checked, since you wouldn't think to look there unless you have seen it happen. The black plastic/phenolic sockets and the brown wafer sockets are the bad guys.
I have never seen a white CERAMIC (not plastic) socket fail, including the new Chinese stuff. They can be made to arc, but they don't carbonize and burn leading to shorts.
After cutting the plate wires I still measure the same resistance across the primary taps.
I measured zero continuity between the plate sockets and chassis.
Does the circuit give you any indication of the required OT for replacement? I know the secondary is 8 ohm, so i'm thinking 15W Tweed deluxe, 8000ohm primary, or 20W Deluxe with 6600 ohm primary. go 20W to be safe?
I measured zero continuity between the plate sockets and chassis.
Does the circuit give you any indication of the required OT for replacement? I know the secondary is 8 ohm, so i'm thinking 15W Tweed deluxe, 8000ohm primary, or 20W Deluxe with 6600 ohm primary. go 20W to be safe?
If you liked the tone of the amp (or what you got to hear of it) try to find an OPT that is about the same physical size and weight as the original. Much of the amps "tone" is determined by how hard the original OPT was pushed toward (or into) saturation.
Fender tended toward undersized OPT's that were near saturation. This often leaded to disastrous results is someone played bass through say a Bandmaster at full crank for a while.
Marshall tended toward oversised OPT's and excess preamp gain for the classic screaming sound.
Fender tended toward undersized OPT's that were near saturation. This often leaded to disastrous results is someone played bass through say a Bandmaster at full crank for a while.
Marshall tended toward oversised OPT's and excess preamp gain for the classic screaming sound.
Thanks for the excellent points.
What do you guys think about the excessive hiss that is produced even though nothing is plugged in? The hiss goes away if the volumes are off, or it is greatly reduced if the tone knob is completely rolled off. I noticed one of the plate resistors is way off (stock is 250k, one is reading 500k+), in general should I be replacing the caps and resistors around the preamp tube to help eliminate the noise?
What do you guys think about the excessive hiss that is produced even though nothing is plugged in? The hiss goes away if the volumes are off, or it is greatly reduced if the tone knob is completely rolled off. I noticed one of the plate resistors is way off (stock is 250k, one is reading 500k+), in general should I be replacing the caps and resistors around the preamp tube to help eliminate the noise?
The two main causes of hiss are resistors and tubes, in that order. It is also possible that his can be caused by an oscillation, which could be due to a bad capacitor or sloppy wiring.
A resistor measuring twice its marked value should be replaced. I would replace the plate load resistors first, and any others that carry current.
A resistor measuring twice its marked value should be replaced. I would replace the plate load resistors first, and any others that carry current.
By FAR the most common fault in valve amps are the anode loads of the triodes - they go high and eventually O/C. One reading twice it's value is 100% faulty and should be replaced ASAP - I'd also advise replacing the others while you're doing, as it's likely they are faulty (or on their way to faulty) as well. Such faulty resistors are a prime source of noise!!.
Use decent quality resistors, 1W metal film would be a good idea, and should prevent them ever failing again.
Historically capacitor problems were mostly leaky coupling capacitors, making the grid of the next valve positive - but anything remotely 'modern' shouldn't suffer from that.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.