RCA BA-6A from scratch

[DaveP]

Letterbeacon,

What does it excel on?

Well this is my situation, I only have 2 evenings a week for electronics and recording and the weather plays a big part, if it rains the noise on the roof means I can't record, likewise if it gets too cold.  So when I can record I don't want to waste time trying this or that, I need to get on with the creative process with familiar tools.  So the short answer is I haven't had time to try it out thoroughly yet, but in the new year I will have time.  What I can tell you is that it needs to be on for 30 mins and balance checked before use, you can't just switch it on like an LA-2A.

Just when I was about to start putting it through its paces, the EMI mod became available and that was too good an opportunity to miss, that's the way it goes unfortunately.

best
DaveP

 

[letterbeacon]

That's fair enough!  I'm kind of building this unit 'blind' (or should that be deaf??) as I've never heard one before.  I'm building it because I've read so much about it (all good things) and I've kind of become slightly obsessed with equipment of this era (especially RCA stuff).  I'll be eager to hear yours in the new year!

 

[emrr]

Who is selling 'matched' 6SK7's?  I've yet to see anyone selling reliably matched small signal tubes.  The RCA factory matching device is fairly complicated, and I have a hard time imagining anyone spending the time to duplicate it.  These tubes drift a good bit in the initial weeks of use, which would explain some of the involved metering options.  If any sellers are truly burning in tubes at full voltage and max current for 100 hours before matching, I'd be interested to know. 

 

[Biasrocks]

Try a BA6A on acoustic guitar, instant Jimmy Page. Loved mine on Bass Gtr and Vocals as well.

I know some folks strap a pair across the stereo bus as well.

Regards,
Mark

 

[letterbeacon]

Who is selling 'matched' 6SK7's?  I've yet to see anyone selling reliably matched small signal tubes.  The RCA factory matching device is fairly complicated, and I have a hard time imagining anyone spending the time to duplicate it.  These tubes drift a good bit in the initial weeks of use, which would explain some of the involved metering options.  If any sellers are truly burning in tubes at full voltage and max current for 100 hours before matching, I'd be interested to know.

That's interesting to know.  I bought mine from an eBay seller that advertised them as 'matched'.  Probably not the most reliable of places to buy them!

Try a BA6A on acoustic guitar, instant Jimmy Page. Loved mine on Bass Gtr and Vocals as well.

Great, thanks for that.  Do you have any samples at all?

 

[lassoharp]

I would only trust matching if the seller had a BA-6 on bench and could verify a constant voltage difference through the entire range of GR.  We of course know that sellers aren't prepared to do this.

Stated another way,  I've had pairs of SK7s and 1612s that show ideal static balance (meaning they probably appear 'matched' on a tube tester) but go too far out of balance under GR, and others where static balance looked relatively lopsided but behaved more closely in the critical areas.  It's hard to tell until you stick them in the real circuit.

 

[letterbeacon]

Hi Dave,

I was going to buy a 5H choke for my BA-6A (as per the schematic) but then earlier in the thread you mentioned that:

The choke I used is on the limit and rather too much ripple comes through for my liking.

I think you said you used a 10H choke.  If you were starting again, what inductance value would you use for the choke?

 

[DaveP]

Letterbeacon,

It was not the lack of Henries that was the problem, it was the lack of current carrying capacity.  My choke was only rated at 100mA and as I said, it was drawing about 120mA, so the choke was over its linear section and operating outside its limits.

best
daveP

 

[letterbeacon]

Thanks Dave -mine's rated at 150mA, which should be enough.

Another quick question:  Can you remember what diodes you used for CR1 (the rectifier for the meter)?  It looks like some sort of zener bridge rectifier - 2.25v reverse voltage with a dc forward current of 10mA.

Thank you!

 

[DaveP]

Letterbeacon,

They are not zeners in the bridge.  The output is AC, probably could hit 70V so you need a 200V diode (for safety) with low forward voltage drop, comparable with what they had in the 50's.  The originals were probably copper oxide or germanium but I used 200V MUR120 part no. 625-5329 from RS.
For lower voltage applications I've used Schottky diodes which have very low Vf drops.  It's always worth checking diode specs for the Vf at low currents, unfortunately there is always a trade-off between Vmax and Vf.
best
DaveP

 

[emrr]

Copper oxide would be correct. 

 

[letterbeacon]

After having been sidetracked by work, I am picking up this project again.

Dave - I'm thinking of copying your heater power supply circuit for my clone and have my heaters use 6.3vdc.  Having built it, would you change anything at all?  I suppose it's a case of building it and then tweaking the dropping resistors until you get 6.3vdc.

One question though - you've specified a 250mA fuse but I've added up the current draw of the tubes as follows:

6SK7 = .3A
6SK7 = .3A
6J7 = .3A
6J7 = .3A
6V6 = .45A
6V6 = .45A
6H6 = .3A

Total = 2.4A

Surely that should be a 2.5A fuse in there, or am I missing something?

Thanks a lot.

 

[DaveP]

Hi Letterbeacon,

I have various low value resistors that I use to get the 6.3V then I buy the nearest Al clad resistor and fit that.

The fuses are always on the mains primary side so they are smaller, that way they last longer cos they don't get so hot, and its safer because the supply is cut off if the TX blows.

Would I change anything?

I would probably not bother with the the meter voltage measuring set-up.  In a radio station with it on 24/7 it would be more prone to drift, but in a studio set-up it has an easier life.

Also I will be using the Mosfet HT delay regulator(see Bridge threads) in future so there will be only one mains switch.
So won't need a separate Heater switch and fuse.  The regulator means the choke is not required either.  The choke radiates hum so losing that is a bonus.

best
DaveP

 

[letterbeacon]

Thanks Dave.

I was reading over on the Valve Wizard website that DC heater supplies must always be regulated otherwise they can be noisier than AC supplies.  Have you noticed any noise problems at all?

Also, I'm having a hard time tracking down a 100r and a 10k 4W muliturn for the Balance A and Balance B pots.  Where did you get yours from?  RS are telling me £93!

I also would like to make sure I understand the attack and release mods you've made:
On the balance/ limit switch, the Single and Double settings are as per the schematic.  When the switch is turned to the 6th position,  you can use R55 to control the attack and release.

R59 is a threshold mod allowing you to set the threshold of the limiter.

Is that correct?

Thank you!

 

[lassoharp]

Also, I'm having a hard time tracking down a 100r and a 10k 4W muliturn for the Balance A and Balance B pots.  Where did you get yours from?  RS are telling me £93!

The balance circuit I built on the breadboard worked great with standard 100R and 10K pots. Lots of range to work with.

 

[letterbeacon]

That's good to know, but it's more the power rating - the BOM specifies 4W.  Although on page two of this thread you link to an eBay listing and I've just noticed that the pot listed is 2W.  Do you think 2W would be ok?

 

[DaveP]

Letterbeacon,

Alan is right, you don't need fancy pots for this job, to put your mind at rest calculate the watts dissipated using the voltages on my schematic and ohms law.  You can get RS ww for about £5 but they can be a bit scratchy, I use Bournes conductive plastic wherever possible and the ceramic version if there is serious current through the pot.  The meter movement on the balance setting is so small that you will be twiddling a multiturn all night to see a change.

Re the DC heater supply:  My understanding is that the diode capacitance can cause some RF that the electrolytics will miss, solution, bypass with 100nF.  Think about it, how can 6.3V AC under the cathode have less hum than DC, maybe balanced centre tapped might, but I would not put money on it.  The GE data sheet for a 6CA7 says, "not suitable for low level audio work unless DC heater is used".  I use a 22,000uF for the first cap and a 47,000uF for the 2nd, only hum I ever heard was from the HT supply through unbalanced tubes.

Looking at your layout on the other post, the choke looks a bit small, are you sure it can handle the current?  If it saturates you will get 100Hz spikes on your HT line.  You may need to screen that choke or hide it behind a PT.

On my schematics, R50 is the threshold, R59 is the attack and R55 is the variable release when in parallel with R53/C20.

best
DaveP

 

[letterbeacon]

Thanks Dave!

Looking at your layout on the other post, the choke looks a bit small, are you sure it can handle the current?  If it saturates you will get 100Hz spikes on your HT line.  You may need to screen that choke or hide it behind a PT.

The choke is rated at 150mA and I think you said that yours draws about 120mA, so I should hopefully be ok.  As for screening - is that a case of putting a steel box around it?

On my schematics, R50 is the threshold, R59 is the attack and R55 is the variable release when in parallel with R53/C20.

Ah yes, I see R50 now.  I see you've added a 100nf (C16) and a 22uf (C16) to the circuit.  Could you explain why you added those there please?

Thanks.

 

[DaveP]

Letterbeacon,

C15 100nF is to bypass the RF to earth that you get from gas regulators (see data sheets) same thing applies to zeners.

C16 22uF is typical of lots of threshold supplies, it helps to keep it steady regardless of whats happening to the HT, it is also additional smoothing.

A steel box will screen the choke as will a steel sheet but the ends of a steel sheet will radiate so a box is best.  Magnetism is only conducted by steel, not blocked as you might assume.

best
DaveP

 

[letterbeacon]

Hi Dave,

Earlier in the thread you said:

I couldn't afford the wafer switches so my switching is in stages.

But I think I've found some affordable switches that do the job.  Because I've found them fairly easily, I'm worried that I'm missing something!

For the Bal/ Limit switch I've found a 2 layer, 2 pole, 6 position rotary switch:
http://www.banzaimusic.com/Rotary-Switch-2-level-2x2x6.html

For the Meter switch I've found a 3 layer, 1 pole, 12 position rotary switch:
http://www.banzaimusic.com/Rotary-Switch-3-level-3x1x12.html

Would these work?  I think they would, but as I say, I'm worried I'm missing something here.