KT66SE B+ problem

[AdamK]

Hi Guys,

Scratching my head here, hoping for some festive support and Christmas cheer.

Attached is power supply schematic for my KT66SE amp.

For some reason, I have a very, very low B+, using my variac at around half uk mains voltage, 140v, my B+ is only around 5v (!)

At  ac 140v, the plates into the rectifier measure around 160v ac, which is what you would expect.

The heater, also, measures OK, 3v ac or so.

Ordinarily, 140v ac is enough to get the amplifier going and B+ would be 270v dc, but no, it's 5v, with both plate and heater voltages into the rectifier being OK.

I did short the amplifier through my variac previously (I installed a meter on the variac and one of the cables detached and caused a short), so assumed I had shorted the rectifier and so I replaced it.

This is when is gets a bit strange.

For two tests, and two new rectifiers, the first switch on everything works OK at 140v, then when I increase ac voltage, the dc disappears.

I have no idea why it would do this, so:

1) Shorted my variac and damaged the amp.
2) Replaced rectifier and first test to 140v was OK, increased ac voltage then B+ reduced from 270 down to 30v.
3) Replaced rectifier again , first test to 140v was OK, switched off then left the amplifier till the next day.
4) Next day, tested again at 140v ac, and this time, despite being ok day previous, B+ was only 5v.

Any ideas?

[AlNewman]

Do you have a lightbulb limiter?

Are you testing with tubes in?

First tests I would do would be with the amp off, discharged, and tubes out.  Test all resistances between all coils of the transformers, and resistances from the B+ rail to ground.  Go through and look for any burnt resistors, bulging or cracked caps, bad solder joints, etc.

[SEL49]

I don't think you have a problem. Maybe your 5U4 filaments are not heating enough for proper reliable tube conduction. I suggest removing the output tubes, removing the variac, plug straight into the mains, and re-evaluate.

If you must have a variable B+ then replace the 5U4 with silicon rectifiers. Then the B+ will reliably change as you turn the variac dial.

[AdamK]

Do you have a lightbulb limiter?

Are you testing with tubes in?

First tests I would do would be with the amp off, discharged, and tubes out.  Test all resistances between all coils of the transformers, and resistances from the B+ rail to ground.  Go through and look for any burnt resistors, bulging or cracked caps, bad solder joints, etc.

Thanks for your reply,

No lightbulb limiter sorry.

With the amp off, discharged and tubes out:

Mains transformer coils:

1) plate to plate (pins 4 & 6) - 49R
2) plate to ground - 23R
3) heater pins (2 & 8) - 1.2R (0R on second meter)

B+ to ground - 10.81k

[AdamK]

I don't think you have a problem. Maybe your 5U4 filaments are not heating enough for proper reliable tube conduction. I suggest removing the output tubes, removing the variac, plug straight into the mains, and re-evaluate.

If you must have a variable B+ then replace the 5U4 with silicon rectifiers. Then the B+ will reliably change as you turn the variac dial.

Thanks for reply.

I think, maybe, my transformer heater is not working, maybe, not sure. 0R on 1 meter amd only 1.2R on another. This is measuring pins 2 & 8.

[SEL49]

I think, maybe, my transformer heater is not working, maybe, not sure. 0R on 1 meter amd only 1.2R on another. This is measuring pins 2 & 8.

Nothing wrong with that. 5V and 3A implies 1.6Ω.

[AdamK]

[/quote]Nothing wrong with that. 5V and 3A implies 1.6Ω.
[/quote]

OK, so, we can rule out the mains transformer?

Its odd that it was working fine up to 200v and then collapsed, the heaters had time, even the KT66'S were conducting.

I do have some BY255 diodes and some high voltage 10nf caps, I'll wire a couple in series and try those but I will have to use a variac as I won't get the 40v or so of voltage drop using a 5u4g.

I can't see it working though as the 5U4G was brand new, if it does work then that would suggest the 5U4G (two of them) were faulty. Or, the transformer heaters, under load, are not working correctly, even though resistance measure OK when it's off.

I've checked for scorched resistors and bulbous caps etc and everything looks OK.

[SEL49]

I can't see it working though as the 5U4G was brand new, if it does work then that would suggest the 5U4G (two of them) were faulty.

It's unfair to suggest the 5U4s are faulty when you only supply half the proper filament voltage. Give them a full 5VAC on the filaments and everything will work as designed.

[AdamK]

I can't see it working though as the 5U4G was brand new, if it does work then that would suggest the 5U4G (two of them) were faulty.

It's unfair to suggest the 5U4s are faulty when you only supply half the proper filament voltage. Give them a full 5VAC on the filaments and everything will work as designed.

OK, I'll try tomorrow, along with the using  diodes for rectification.

I've attached my previous test readings sheet. Over the years, many times I've tested the amp at 140v to start with and never had any conduction issues.

Thanks for your help, I hope you read my findings tomorrow

[pdf64]

A lightbulb limiter is required.

[Willabe]

A lightbulb limiter is required.

Look in here to build 1.

https://www.sluckeyamps.com/misc/Amp_Scrapbook.pdf

[Willabe]

For some reason, I have a very, very low B+, using my variac at around half uk mains voltage, 140v, my B+ is only around 5v (!)

At  ac 140v, the plates into the rectifier measure around 160v ac, which is what you would expect.

heater, also, measures OK, 3v ac or so.

Ordinarily, 140v ac is enough to get the amplifier going and B+ would be 270v dc, but no, it's 5v, with both plate and heater voltages into the rectifier being OK.

Why are you doing that?

Stop testing the amp running it at 1/2 the input acv.

You are only getting ~1/2 the heater acv the heaters need to heat up enough to release the electrons from the cathode (K) sleeves coating.

You'll greatly shorten the tubes life or kill the tubes doing that. And you can't get any measurements that mean anything that way.

Use a light bulb limiter. When the bulb shows you the amp has no shorts, then unplug the LBL and put it away.

[AlNewman]

B+ to ground - 10.81k

That looks questionable, based on your schematic.  Did you give it time to charge?
Build an LBL.  Best $10 you'll ever spend.

[AdamK]

For some reason, I have a very, very low B+, using my variac at around half uk mains voltage, 140v, my B+ is only around 5v (!)

At  ac 140v, the plates into the rectifier measure around 160v ac, which is what you would expect.

heater, also, measures OK, 3v ac or so.

Ordinarily, 140v ac is enough to get the amplifier going and B+ would be 270v dc, but no, it's 5v, with both plate and heater voltages into the rectifier being OK.

Why are you doing that?

Stop testing the amp running it at 1/2 the input acv.

You are only getting ~1/2 the heater acv the heaters need to heat up enough to release the electrons from the cathode (K) sleeves coating.

You'll greatly shorten the tubes life or kill the tubes doing that. And you can't get any measurements that mean anything that way.

Use a light bulb limiter. When the bulb shows you the amp has no shorts, then unplug the LBL and put it away.

I thought this was standard way to test an amplifier and, if not, why have a variac, its only to test briefly and as stated it worked perfect before.

But, I will make a little LBL and follow your advice as it is very much appreciated could I ask?

1) would a 100w bulb be OK, only that I have some.
2) I suspect the bulb will stay bright, assuming it does, how will I know if its the PT or supply caps that are shorted, or, does it just indicate there is a short.
3) will take a week or so to make due to festivities, I hope you will all continue to read when I post again. I was really hoping to have this resolved for Christmas but not to be.

Merry Christmas 🙏

[Willabe]

You are only getting ~1/2 the heater acv the heaters need to heat up enough to release the electrons from the cathode (K) sleeves coating.

You'll greatly shorten the tubes life or kill the tubes doing that. And you can't get any measurements that mean anything that way.

Yes, 100w bulb is very good for this sized amp.

I'll write more later, it's Christmas eve and I have a lot to do.

[Willabe]

I thought this was standard way to test an amplifier and, if not, why have a variac, its only to test briefly and as stated it worked perfect before.

It is not a standard way to test an amp. All those test numbers you wrote down taking voltage readings with different acv feeding the amp mean nothing. You don't need them for anything at all.

The only voltage readings that will mean something are when the amp gets full acv.

A variac can be used for a number of things but what you are doing with it is not what it is for.

What do you think all those readings are showing you?

2) I suspect the bulb will stay bright, assuming it does, how will I know if its the PT or supply caps that are shorted, or, does it just indicate there is a short.

Yes, if the bulb stays bright, it means there's a short. What the short is you then have to find it.

You'll cross that bridge if/when you get to it. 

[AdamK]

What do you think all those readings are showing you?



Yes, if the bulb stays bright, it means there's a short. What the short is you then have to find it.

You'll cross that bridge if/when you get to it.

As a novice I do appreciate all your comments and advice, especially from a global moderator. I will inevitably do things that will appear pointless to an expert.

The readings chart was simply a step by step set of measurements I took, rightly or wrongly, and use them as a reference.

This time round, measurement taken at 140v, 160v and 180v were exactly the same as last year. When I got to 200v, B+ collapsed as described above. So, at least it showed me that.

I'll post again with the LBL test and I hope you can take the time to read and respond.

[Willabe]

I'm not an expert. 

The readings chart was simply a step by step set of measurements I took, rightly or wrongly, and use them as a reference.

You don't need a 'reference' for what the tube dcv's are with the amp's input acv way below normal. It's not a 'reference' for anything. That's not telling you anything because that's not what the amp was designed to see at it's power transformers (PT) primary input.

To check the amps PT secondary acv's and the tubes dcv's you measure them at full normal wall acv. Then you can see if the voltages are right. There is no 'step by step set of measurements' that you should be taking. It's a total waste of time and can damage the tubes/amp.

You go by the acv/dcv listed on the amps schematic or if it's a home made amp, you go by the acv/dcv readings you took and wrote down when you finished the amp build. 

And if the amp is working fine, you don't need to check any voltages. Because there's nothing wrong.

This time round, measurement taken at 140v, 160v and 180v were exactly the same as last year. When I got to 200v, B+ collapsed as described above. So, at least it showed me that.

No, it didn't show you anything because the amp doesn't run at those acv/dcv's. 

At the correct voltages the amp was designed for, it will put the correct amount of load and stress on the PT, tubes and all the other parts.