5w Screen Grid Resistors solder melt

[Pryde]

Hello folks

I am working on a Traynor YBA-1
Schematic attached (page 8 closest version)

The issue is off of each EL84 screen grid (pin 4) there are 1k 5w resistors that apparently are getting so hot the solder is melting in the eylet hole and disconnecting the wire (leading to pin 4)

Visual inspection shows no burn marks or browning at all on either of them. I checked resistance and one is 895 ohm, the other is 920 ohm. so there is some drift from 1k but is that causing the issue?

I don't want to just resolder as they likely will melt it again over time.

Also, the schematics all have pin 4 of both tubes connecting to 1 10w 470 ohm resistor but this one has individual 1k to each pin.

Thanks for your help on how to proceed

[soundguruman]

Hello folks

I am working on a Traynor YBA-1
Schematic attached (page 8 closest version)

The issue is off of each EL84 screen grid (pin 4) there are 1k 5w resistors that apparently are getting so hot the solder is melting in the eylet hole and disconnecting the wire (leading to pin 4)

Visual inspection shows no burn marks or browning at all on either of them. I checked resistance and one is 895 ohm, the other is 920 ohm. so there is some drift from 1k but is that causing the issue?

I don't want to just resolder as they likely will melt it again over time.

Also, the schematics all have pin 4 of both tubes connecting to 1 10w 470 ohm resistor but this one has individual 1k to each pin.

Thanks for your help on how to proceed

a pair of 1K resistors is expected with el34, the screen grid is needing current limiting.
The melt point of the solder has a lot to do with it
many solder would not melt under those circumstances
unless the bias was way too hot maybe...
The solder I use would not melt, but there are low temperature solders too.
So it's a good time to check the melt point, after all the old solder can be removed and replaced.
The replacement solder can have a higher melt point.
It's probably the bias was too hot or lower temperature solder was used.
It probably took some years for the solder to deteriorate to that point.

so re-soldering may last several years with better solder

[Chuck H]

I'm guessing you meant EL34 tubes, not EL84.?. I can't see anything about the info or schematic that should cause excessive screen dissipation. Take a voltage reading across the screen resistors and calculate the dissipation.

Ant other readings could also help. Bias? Plate voltages? Screen voltages? Are the tubes glowing red other than the filaments? Is the correct fuse in place (or circuit breaker)?

[Chuck H]

Do you really think the suggestion of solder with a higher melting point is sage for a screen grid circuit that's melting the joints? Yes, re-solder them. Take some readings to determine what's going on too though.

[Pryde]

I'm guessing you meant EL34 tubes, not EL84.?. I can't see anything about the info or schematic that should cause excessive screen dissipation. Take a voltage reading across the screen resistors and calculate the dissipation.

Ant other readings could also help. Bias? Plate voltages? Screen voltages? Are the tubes glowing red other than the filaments? Is the correct fuse in place (or circuit breaker)?

Yes EL84 was a typo. I meant EL34s. It appears the Xicon 5w resistors look fairly new in this amp and it is possible whoever put them in ethier: a) forgot to solder them at one end, or b) did a very poor job with little solder in the joint.

Anyway, I will resolder properly and take some measurements and report back tomorrow. thanks everyone!

[Pryde]

OK an update with some numbers. I resoldered the offending joints and played monitored the amp for 45 minutes with no issues (1k 5w screen resistors are not hot at all)

Plate voltage: 407vDC
bias current: 48 m/a. (running about 76% MPD for the EL34's)
Screen grid voltage 420vDC
Control grid -31.6vDC

These numbers help steady during test period.

Issues or just bad soldering previously?

[g1]

The measurement you need is the voltage across each of the screen resistors, from which you can calculate the screen current for each tube.
Also, the screen voltage should be less than the plate voltage, not higher. Check that the screen feed is coming off the proper node of the power supply,

[Pryde]

The measurement you need is the voltage across each of the screen resistors, from which you can calculate the screen current for each tube.
Also, the screen voltage should be less than the plate voltage, not higher. Check that the screen feed is coming off the proper node of the power supply,

OK I attached a pic of the 1k 5 watt screen resistors for reference. The bottom of each one is connected to pin 4 of each tube. You can see where I just freshly soldered them.

ok
The left one measures -4.2vdc voltage across
right one measures -5.7vdc

As far as higher screen vs plate voltage and nodes, the top eyelet (where the 3 resistors connect) a wire runs from that straight to one of the filter caps.

Not sure how/what I would do to make screen voltage lower than plate?

[g1]

Those voltages sound ok on the screen resistors. No reason for them to overheat unless tubes are failing when passing signal.

As far as higher screen vs plate voltage and nodes, the top eyelet (where the 3 resistors connect) a wire runs from that straight to one of the filter caps.

That filter cap should be connected to a 470 ohm 10W resistor. The other end of the resistor should go to the first filter cap, the one that feeds the output transformer. See the schematic. It's possible the resistor has been removed (bypassed).

[Pryde]

I do see what you are saying and there is certainly no 470 ohm 10w resistor to be found anywhere. But the schematic also does not show these 2 5w screen resistors either?

Would you suggest that I put a 10w 470 ohm in series between the filter cap and that top eyelet connection while keeping the 5w 1k resistors in there?

[g1]

It's hard to say without knowing how the filter caps are connected. Yes, it should be installed but it really should be between the filter caps as shown in the schematic.
The 2 1K's should stay in, they are replacing the single 470 ohm shown below the power tubes.

[Chuck H]

Make sure the plates are getting their HV from the first filter node on one side of the choke and the screens are connected to a second filter node, down stream on the other side of the choke.

[g1]

Chuck, from the schem. mentioned, this should be the version without the choke. The 470R 10W that seems to be missing should be in the same position. Unless someone replaced it with a choke .
More pics of the filter cap area would help.

[Pryde]

Here are a couple photos of the caps. recent F&T can caps 50/50.

No resistors anywhere.

The left cap is tied to rectifier and red/yellow from choke (I believe)
Right cap is tied to those screen resistors and a black wire from choke.

happy to post more photos if needed to help me sort this

[g1]

The schematic on page 6 of that pdf shows the version with a choke.
What voltage drop do you measure across the choke?