Kit 6550 resistor failures and voltage irregularities

[Revoli]

First a huge thanks to those reading this and offering advice

I have a persistent problem with my Kit6550 amp. Before I dive in and do stuff could someone check my sanity on what planning to try??? Amp ran for many years no problems. Now symptoms as follows:

(with reference to original circuit diagrams)

Resistors R35,36,37,38 (560ohm 7W Large) overheating and failing (replaced twice).

Test Voltages are off in two cases:
a. between V4 pin 8 8.8V now versus 6.7V originally on build
b. between V3 pin 8 8.5V now versus 6.7V originally on build
c. Valves 5,6,7,8 between pins 2 and 7 now 2.6v versus 5.6V on original build

I replaced most caps and many resistors when this first happened as amp getting old and whilst in there seemed like a good idea (and suggested here). Now initial thoughts were to:

1. Check all wiring is still sound
2. Up resistors R35,36,37,38 to from 560Ohm 7W (https://uk.rs-online.com/web/p/throu...istors/2061035) to 560Ohm 14W wirewound resistors (https://www.cricklewoodelectronics.com/14W560R.html) but am wary it all worked fine on 7W resistors previously so whats changed and am I just shifting problem elsewhere???
3. Should I replace all valves as a precaution (is this good practice to periodically do before thay fail???)

I would attach circuit diagrams and the testing details for ease of reference (also has my full test results scrawled on them etc) but files too big for forum. If its helpful I can email!

Thanks in advance.

[John Caswell]

Hi,
Well if the resistors are overheating that means the valves are drawing too much current, and changing the resistors to 14Watt types is shifting the problem elsewhere.
Sooo... either you have a set of knackered valves (how long have they been installed?) or the coupling capacitors to the 6550 grids are leaky.
You can check this by measuring on the 6550 grids (Pin5) to chassis. It should read 0V or very very low. If it does not read low, power down, pull out the 6550 power up and measure again. If the voltage is now very low your 6550 are shot, if it still measures high then I would suspect leaky coupling capacitors (C11,12,13,14)
The voltages on pins 8 of the 6550s should all read somewhere in the region of 44Vdc
Voltages across pins 2&7 of the 6550 (V5,6,7,8) should read approx 6.3Vac and across pins 4&5 of the ECF80 (V3,4) approx 6.3 Vdc and across pins 2&8 of the 5U4s (V1,2) approx 5Vac.
I usually suggest replacing output valves on an 18month to 2 year base as they are generally run hard.
Have another look and report back here.

John

[Revoli]

Thanks John. A thorough check of wiring threw up discrepancy. Circuit diagram shows a connection for the 6550 valves between pins 1 and 8. My amp never had that (and worked fine). On 6550 valves pin 1 has no connection I believe?? Should I have this connection between pins 1 and 8, I assume it makes no difference?

[A Stuart]

Is this just a typing error saying pin 8 is no connection?
Three out of three sites I have just looked at say pin 6 is the one with no connection.
This site https://frank.pocnet.net/sheets/049/6/6550.pdf describes pin 1 as "Base Sleeve" connection.

EDIT - Sorry, on re-reading your post, ignore the above comment. My best guess is that pin 1 could be connected to 8 just to "earth" it to the cathode, so that it is not floating uncontrolled.

And, don't forget, John pointed out there is something seriously amiss with the heater voltage pins 2-7, assuming 2.6 volts is a true reading and you didn't inadvertently have a range-doubling function on your meter.

[John Caswell]

On a lot of octal valves, especially the metal types eg 6V6 6L6 6SG7 etc and many glass power valves with a metal rim around the octal base, the case and the metal rim were connected to pin 1 which was generally earthed or connected to the cathode.
In the case of the metal valves it was specifically recommended to earth pin1 for obvious reasons, shock hazard and screening.
So pin 1 is really a hangover from those days. Earthing it or connecting to the cathode (generally pin 8) will do absolutely no harm at all. (Just don't connect it to pin 8 and earth at the same)
Pin 6 on KT88/6550 series etc generally has no pin in the valve base so I often use it to fit the low value grid "stopper" between pins 5 and 6..

John

[Revoli]

Once again thanks for replies.

Sooooo....

I have added connection pin 1 to pin 8 on 6550s. Shouldnt have made a difference but its on the wiring plan and better to be safe etc.

Came to test (with valves in) switched on and got the magic dragon signal (that ominous puff of smoke) from rectifier. My rectifier was diode construction (which i think was the original design for this amp). I am wondering if it was on the way out and might have caused my probs?? Anyway will have to replace - one of those new fangled KBU4D rectifiers I suppose would make sense.

Once done would it make sense to test sans valves first?

On the bright side I will have lots to do in the christmas lockdown .....

[John Caswell]

Puff of smoke from heater bridge rectifier implies that either the an ECF80 valve heater is s/c (unlikely) or the output smoothing capacitor is s/c (more likely) A KBU04 is fine
Test without valves first of all to make sure that you have roughly the correct ac and dc voltages at the right points. They will all be high because no current is being drawn.
Then fit the 5U4 rectifier valves and check that you have (high) dc HT. If all is well fit the 6550 and ECF80 power up again and measure all the necessary voltages. The 6550 voltages should be very near to that stated but the ECF 80 voltages are likely to be all over the place - see my FAQ re ECF80.
Once again any problems report back here

John

[Richard]

Just a couple of thoughts,

When the failed diodes etc have been removed, and without valves in, I'd check the 3 Mains TX output AC voltages at the tags where their wires join. They should all read higher than the schematic without load whereas you found low heater voltage which is strange. If needed (reading not correct) disconnect those wires to isolate them from the following resistors and caps in case they are at fault and check again.

I wouldn't use larger W cathode resistors. Far better for them to fail in a fault situation than to continue to allow excess current to be drawn and take out the OPTX primaries for example.

You don't say how old the valves are but after rebuilding change all the valves for known good or new ones.

[bob orbell]

My thoughts on the first post, John suggests C11-12-13 and14 should be checked and I agree if they are paper in oil, but C15-16-17 and 18 have not been mentioned could be the problem. Bob

[Revoli]

Originally Posted by Richard

Just a couple of thoughts,

When the failed diodes etc have been removed, and without valves in, I'd check the 3 Mains TX output AC voltages at the tags where their wires join. They should all read higher than the schematic without load whereas you found low heater voltage which is strange. If needed (reading not correct) disconnect those wires to isolate them from the following resistors and caps in case they are at fault and check again.

I wouldn't use larger W cathode resistors. Far better for them to fail in a fault situation than to continue to allow excess current to be drawn and take out the OPTX primaries for example.

You don't say how old the valves are but after rebuilding change all the valves for known good or new ones.

Thanks all for suggections. I will work through and record here as I address each. May not be in order....

Firstly valves. Too old possibly. I have bought new and will proceed with those once other suggested tests complete.

Tests. Without valves first:

Transformers without diodes/bridge rect (ie removed completely)
mains trans black wires: 380v; and 380v
purple wires: (v2, pin 2) 4.53v; (V2 pin 8 0v)
grey wires: 2.87v; 2.87v

These are lower than schematic but incoming mains measuring 197v which I guess would account for the difference? (mains 82% lower, if adjust numbers by same factor looks better but guess this isnt a linear relationship....?)

Please shout if any of that is of concern.