Planning my BA-6A

[letterbeacon]

I turned up the input and the output to max (with nothing running through it) and I could just hear a tiny amount of hiss. 

I'd call that a success!  10W PP pentode output stage cranked to max and tiny amount of hiss is a high five situation. 

Haha!  I was very pleased with that as I spent a lot of time trying to make this thing as quiet as possible.  I paid particular attention to the lead dress of the AC heaters, as well as kept my signal grounds and drain grounds a separate.  I also went a bit overboard with the shielded cable.

I've just run a random drum loop through it and it sounds great!

And it will over a fairly wide range of voltage variations from what the manual states. 

Congrats!

I didn't have much time to spend with it tonight, but I should have more time tomorrow to really see what it can do. 

All you need to do now is put the mains on the 220V tap and not the 240V and you should be nearer the mark, thats what those taps are for, fine tuning.

Great, I'll give it a go tomorrow.

Thanks very much both of you, as well as Doug and everyone else in the thread for getting me this far!  It's hugely appreciated and I've learnt loads!

 

[buildafriend]

Are there any pictures that I can see? 

 

[letterbeacon]

Well done Matt,

All you need to do now is put the mains on the 220V tap and not the 240V and you should be nearer the mark, thats what those taps are for, fine tuning.
best
DaveP

I connnected up the 220V tap and I'm happy to say my voltages are within range of the RCA voltages (good enough for me anyway, especially as I used a fair amount of NOS CC resistors which aren't hugely precise).  Here is what I measure now:

6SK7
V1 Pin 8 = 145vdc (135)
V2 Pin 8 = 140vdc (135)

V1 Pin 6 = 96vdc (103)
V2 Pin 6 = 96vdc (103)

V1 Pin 5 = 4.79vdc (5.15)
V2 Pin 5 = 4.80vdc (5.15)

6J7
V3 Pin 3 = 69.5vdc (71)
V4 Pin 3 = 43.8vdc (71)

V3 Pin 4 = 36.3vdc (42)
V4 Pin 4 = 36.3vdc (42)

V3 Pin 8 = 1.10vdc (1.18)
V4 Pin 8 = 1.19vdc (1.18)

6V6
V5 Pin 3 = 269vdc (275)
V6 Pin 3 = 269vdc (275)

V5 Pin 4 = 277vdc (285)
V6 Pin 4 = 277vdc (285)

V5 Pin 8 = 14vdc (13.1)
V6 Pin 8 = 15.57vdc (13.1)

6H6
V7 Pin 4 & 8 = 63vdc (66)

0D3
V8 Pin 5 = 146.4 (155)

5R4GY
V9 Pin 4 & 6 = 565vac (315vac)

The 6J7s are still a little unmatched and despite the unit sounding good, I think I'm going to buy a matched pair anyway.  They're only $5 each!

Are there any pictures that I can see? 

There absolutely will be in the build thread!  I just want to get the dials engraved and get some feet for the underside and then I'll upload some pics.

Right, I'm off to play with this and my BA-2A!

 

[letterbeacon]

I've noticed today that my meter doesn't quite go up to 0 (full scale) any more when in GR mode.  It did yesterday, so why not today?  All I've done since yesterday is increase the B+ voltage and balanced the cathode voltages using a DMM.  I've adjusted the zero adjust pot until I can't turn it anymore, but it only gets me as far as 1.

In GR mode the meter is basically measuring the current across V1's cathode resistor, right?

Hmmm, I'm stumped!

 

[lassoharp]

I've noticed today that my meter doesn't quite go up to 0 (full scale) any more when in GR mode.  It did yesterday, so why not today?

Two things come to mind

1) Varying supply voltages - regulated supply is drifting

2) If your 6SK7s are brand new they are undergoing burn in and some irregularities may be encountered until they settle it.  This may apply to the regulator tube but I've never used that one before so can't say firsthand.  In general all of the new tubes including the rectifier may show changes in performance until well burned in.  Outside of voltage drifts - I just saw this happen with pair of 1612s on an SA39 I was restoring - they would go from no thumping to thumping - back to no thumping over the course of several on/off cycles and days.  Doug (emrr) confirmed that that behavior can happen during the burn in phase so I would keep an eye on that too.  Eventually they should settle in to greater consistency.

Lastly, even after burn in tubes can sometimes seem to behave differently on different days.  I've just noticed that over the years with guitar amps and on my SA39B, which has had the same tubes for the past three years, and a regulated supply - yet some days the meter will drift a bit here and there.  The first thought is always changes in line voltage but it's seldom practical to check that every time you see or hear a difference so hard to confirm for sure.

With new 6SK7s you should have ample range for zeroing with the factory adj pot and the mechanical adj on meter face.

 

[DaveP]

Matt,
All of what Alan said is true and in addition:-

Your 6SK7's were unbalanced before and the higher voltage probably gave the right meter reading, now they are both the same it is too low for the meter.  Your cathode voltages are still about 7% lower than spec which is probably enough to keep the meter off 0GR.

The 6SK7's plate voltages are high, which indicates lower voltage drop across the 22k's.  The voltage feeding your interstage is probably too low as are all your voltages generally, so you could probably restore the correct voltage by reducing R41 from 2.2k to 1.8k, this will have no other effect on performance.  You can check this simply by crock clipping a 10k res in parallel across the 2.2k.
best
DaveP

 

[letterbeacon]

Ok, I'll wait a few days and let the tubes burn in before I start to worry.

What does worry me slightly is this:

With new 6SK7s you should have ample range for zeroing with the factory adj pot and the mechanical adj on meter face.

I'm right on the end of both the ba6a adjust pot and the adjust pot on the meter and still I can't zero it.

The 6SK7's plate voltages are high, which indicates lower voltage drop across the 22k's.  The voltage feeding your interstage is probably too low as are all your voltages generally, so you could probably restore the correct voltage by reducing R41 from 2.2k to 1.8k, this will have no other effect on performance.  You can check this simply by crock clipping a 10k res in parallel across the 2.2k.

Good idea, I'll give this a go too.

Thank you!

 

[emrr]

How long is your warmup?  Some limiters take hours for the meter to come all the way up.

 

[letterbeacon]

I left it on for over an hour today, but still the meter didn't reach full scale...

Another thing I've noticed is that if the input is turned up quite high, but the output is completely off, I can hear the signal quite faintly on the output.  I traced the signal and it seems like the output attenuator (the one a lot people use on their 1176 builds) is letting through signal even when it's supposed to be attenuating down to zero, which is slightly weird.

I've also got some crackling on the output now, and it blew a fuse this afternoon!  Not sure it's 100% yet!  I'll get some fresh eyes on it tomorrow.

 

[letterbeacon]

Your 6SK7's were unbalanced before and the higher voltage probably gave the right meter reading, now they are both the same it is too low for the meter.  Your cathode voltages are still about 7% lower than spec which is probably enough to keep the meter off 0GR.

Do you think it's worth me connecting the 200VAC input tap on my PT to raise the voltages further?  My worry then is that the VAC on the 5R4GY anodes will be much higher, and according to the datasheet when it's near 600VAC I need to allow the filaments to heat up 10 seconds beforehand.  Unfortunately I don't have a way of doing this.

 

[letterbeacon]

Another thing I've noticed is that if the input is turned up quite high, but the output is completely off, I can hear the signal quite faintly on the output.  I traced the signal and it seems like the output attenuator (the one a lot people use on their 1176 builds) is letting through signal even when it's supposed to be attenuating down to zero, which is slightly weird.

Out of curiosity, I was looking at the Drip documentation for his PCB build and he specifically mentions not using the Hairball attentuator as it could 'melt'.  I've got a -40dB pad in front my mine using 2W resistors (a la the Gates limiter) which I thought would take most of the power.  Do you think the fact I'm hearing the signal faintly when the input is up high and the output is 'off' is due to the attenuator struggling to handle the power?

 

[DaveP]

Matt,

I would not use the 200V tap.  Better to just raise the voltage on the stage where you have a problem than to stress the whole unit.  You are finding out why we don't use 10% resistors anymore, if you get a combination of high end values it can throw your voltages out.
best
Dave

 

[DaveP]

Quite often transformers and tubes sing a little on a sine wave on full volume as the parts vibrate in sympathy, this is probably all that's happening, unless its getting red hot that is!
best
DaveP

 

[letterbeacon]

Quite often transformers and tubes sing a little on a sine wave on full volume as the parts vibrate in sympathy, this is probably all that's happening, unless its getting red hot that is!
best
DaveP

Ah ok, I thought it might also explain the crackling on the output as well, but maybe not.  I have a Mallory attenuator in my BA-2A which I might swap in just to rule it out.

I had a thought: my voltages are within 10% tolerance of the specs in the RCA manual, and I imagine this would have been good enough for the RCA factory because of the large tolerances on the components available to them back then.  This is why there's a zero adjust on the meter, to take into account the variable voltages they'd get with each build, right?  I'm beginning to think there's something wrong with the meter -maybe I haven't calculated the shunt properly or something.

I can't dial in zero when it's on GR mode
I had to change the resistor in the meter switch from 16k to 6.8k for the needle to line up with the heater mark
Despite the voltages being within 10%, when I select V1 on the meter switch, the needle doesn't read within the 'check' area of the scale.

I think we calculated earlier in the thread that I should use a 2k shunt.  The unit requires a 82r shunt already, so I put an 82r and a 2k in parallel across the + and - of the meter.  Hang on, just typing this out and I've realised something, shouldn't they be in series?

I can't check at the moment, but that could be it...!

 

[DaveP]

Matt,

The fact that the tubes are not coming into the red area either means that maybe the shunt is a little too low in value, so too much is being bypassed through it.  The shunt and the 82 should not be in series, the shunt goes across the + & -.  Try increasing the value of the shunt in stages until you have the meter control back from one end.  i.e. 2k to 2.2k then 2.7k if required.
Hopefully this should sort everything out.
best
DaveP

 

[letterbeacon]

Ok will do, but I don't quite get the theory behind not having them in series:

The BA6A schematic has a shunt of 82r across the 500uA meter (R60).  I need to put a 2K shunt across the my 50uA meter to make it a 500uA meter.  As I understand it then, I need a total of 2820r across the meter.  If I put the 2K in parallel with the 82r I'l only have 78r across the meter.

I'm sure you're right, but I don't quite understand why!

 

[emrr]

my voltages are within 10% tolerance of the specs in the RCA manual, and I imagine this would have been good enough for the RCA factory because of the large tolerances on the components available to them back then.  This is why there's a zero adjust on the meter, to take into account the variable voltages they'd get with each build, right? 

The zero is mostly about adjusting for aging GR tubes, only somewhat for AC variations, etc.  I have an RCA/NBC limiter pair that are impossible to line up at zero.  I rebuilt everything around the metering circuit and power supply with 1% metal film, makes no difference.  I can go through a big pile of tubes and find some that will draw enough to align to zero, and within a month draw will have dropped off enough that it never hits zero.  They still compress well and you adjust your eye to the meter scale.  The scale means very little anyway, it's certainly not a precision gain reduction meter. 

 

[DaveP]

Matt,
look again at the original schematic attached, R60 connects the meter minus to earth.  It is only across an auxillary meter if it is connected.  Your 2k is completely separate as its just to convert a 50uA meter to a 500uA meter.
best
DaveP

 

[letterbeacon]

Matt,
look again at the original schematic attached, R60 connects the meter minus to earth.  It is only across an auxillary meter if it is connected.  Your 2k is completely separate as its just to convert a 50uA meter to a 500uA meter.
best
DaveP

Of course you're right, it does!  Gah!  All those parallel lines confused me!  So at the moment, as I have the 82r and the 2K in parallel, it means my shunt is only 78r.  Taking the 82r out will hopefully bring the meter to what it should be.

The zero is mostly about adjusting for aging GR tubes, only somewhat for AC variations, etc.  I have an RCA/NBC limiter pair that are impossible to line up at zero.  I rebuilt everything around the metering circuit and power supply with 1% metal film, makes no difference.  I can go through a big pile of tubes and find some that will draw enough to align to zero, and within a month draw will have dropped off enough that it never hits zero.  They still compress well and you adjust your eye to the meter scale.  The scale means very little anyway, it's certainly not a precision gain reduction meter.

Thanks Doug, that's good to know.  I guess I'm being a bit pedantic when it comes to getting everything working as it says it should in the RCA manual.  It sounds fantastic on everything I've tried it on so far!

 

[DaveP]

I have only ever wired up a circuit perfectly once or twice, there is nearly always a silly mistake.

To avoid those, I use a spare schematic and as I wire up a component, I go over the schematic lines in red until its all red.  Sometimes even that is not enough to counter my own stupidity! :
best
DaveP