RCA Ba2A Pre amp qestions & answers

[gary o]

Interesting info.....I enjoy DIY ing & usually learn the hard way trial and error....Im starting to understand earthing & things like that....The pres i have DIYed have a separate PSU in a box 6 feet away...

My question is will carefully laying out the components on turrets make much difference if there is no nearby power supply part like the power transformer...

does the layout keep noise lower in general ?

In the past in my ignorance & lack of info or pictures I have built fairly complex tube gear just bird nesting everything the shortest route I could see & then tried different earth paths to hear which is best ...it usually ended up looking like the star ground diagram it was nice too see that my ears led me to the star grounding before I came across the star grounding info on the web

For instance my ghetto version of the BA2A is just 2 tube sockets close together with components soldered between the 2.... the whole powers supply is elswhere in a metal case

So Im wondering if a better layout would make lower noise in my circuits, not that they sound noisy at the moment but if my messy layout may be noisier it could be time for a tidy up....maybe I need to measure what noise i do have....

Looking forward to seeing letterbeacons BA2A & hearing his reports on the sound......

Thanks for all info here am learning lots.....

 

[letterbeacon]

I don't think turret board layouts are necessarily quieter at all.  My choice was purely because I had never made a turret board before and wanted to have a go.  In fact, I wish I had gone point to point with this one because the turret boards are taking up a fair bit of space inside my fairly small enclosure.

I can't wait to hear my BA-2 either!  Just waiting on the transformers now!

 

[emrr]

Back on page 4:

http://www.groupdiy.com/index.php?topic=26675.msg363779#msg363779

 

[DaveP]

Gary,

Sounds like shortest route between components is a good way to go as long as the earthing is correct, remote power supply is never a bad idea.

Laying components on turrets does look neat and is handy for removal for servicing, but it can often mean longer wiring runs as a result, a kind of trade-off of neatness for noise if its not done carefully.

Re, measuring noise, I'd welcome some input on that myself.  Just today I've been measuring some triode wired RF pentodes to find some low noise types, but my scope only goes down to 1mV/div and you can just about see the difference between 100uv and 200uV on that.  Steve Bench reported figures of 0.67 uV in his tests.

best
DaveP

 

[gary o]

Thanks dave yeah the turret board designs usually make the signal path longer so more chance to pick up something un wanted l...like and ariel i guess & therefore more important to layout sensibly....I nearly put that in my previous post...

I read somewhere about tubes can be too close to each other too.....

sadly I have no noise measuring gear & wouldnt know how to use it really.....for now I been using some RME software, basically a huge sensitive meter, I can see the noise on the meter & adjust the circuit to watch the level go down hopefully its been handy to tune my gear....

thanks once again....

 

[emrr]

BA-2 layout

BA-2 guts

 

[letterbeacon]

Great, thanks for that Doug, very interesting!

 

[letterbeacon]

So I've finally got this thing powered up and I have a few questions regarding the power supply.

Here are some of voltages I should be expecting:

V3 / 215AC-0-215AC according to voltage chart, differs from schematic 230-0-230 / 288VDC
V2 plate 232VDC cathode 8.8VDC
V1 plate 73VDC cathode 2.4VDC

I'm using a Hammond 370x for the PT which is 240-0-240 rather than the 230-0-230 specified by the original schematic, so as discussed earlier in the thread, I'm going to have to change R8 & R9 to get the voltages in spec.

The voltage at the plate of the 6X5 is 288VDC so, sing Ohm's law, the voltage after the R9 (ie. the voltage that goes to pin 1 of the OT) is 248VDC.  The current draw of the amp is 5.6ma x 7.2k (which is the combined resistance of R8 & R9) = 40v voltage drop.  288 - 40 = 248VDC.

As I'm just getting my head around all this, could someone confirm that is correct please?

Thank you.

 

[lassoharp]

The voltage at the plate of the 6X5 is 288VDC so, sing Ohm's law, the voltage after the R9 (ie. the voltage that goes to pin 1 of the OT) is 248VDC.  The current draw of the amp is 5.6ma x 7.2k (which is the combined resistance of R8 & R9) = 40v voltage drop.  288 - 40 = 248VDC.

Yes, this is correct.


ps - (probably a typo but technically you would mean the cathode of the 6X5)

 

[letterbeacon]

Great, thanks lassoharp.

So I've powered up the unit and have the following voltages with the schematic's voltages in brackets:

v3 plate = 295 (288)

after r8(4.7k) & r9(3.9k) = 247 (248)

v2 plate = 211 (232)
v2 cathode = 7.83 (8.8)

v1 plate = 80 (73)
v1 cathode = 2.3 (2.4)

So none of them are exactly right, but are in the right ball park.  I'm slightly worried about V2's voltages, are the too far out do you think? Apart from R7 R8 & r9 all the resistors are carbon comps which might go some way to explain the discrepancies.

I plugged a mic in and it works!  Kind of.  I can hear myself but there is a very loud buzz, which sounds like a grounding problem.  I laid out the ground schemes as per that Jensen diagram that was posted in the other thread but I'll double check them.

I haven't done anything with the 6.3VAC CT at the moment - shall I ground this?

The Hammond 370x PT is also very hot.  I can touch it for a couple of seconds before it gets too hot to touch.  Is this normal?  It's the first time I've used one so I have no idea.

emrr (who sold me the input transformer) recommended that:

On the secondary [of the input trasnformer], try connection both ways when you have test gear, and see which has the better top end response.  Usually one way does, due to capacitive distribution within the coil.  Once you arrive at the better connection, then flip-flop input terminals as needed for correct polarity through unit. 

I don't have much in the way to test frequency response, is there any software that you can recommend?  How about RMAA?
Thanks for all your help so far!

 

[emrr]

The voltages you show are fine; your output is the higher primary resistance type, and that will give a lower voltage like you are seeing.  The 6.3CT is part of your loud buzz.  The Hammond probably shouldn't get that hot. 

 

[letterbeacon]

Thanks Doug, I'll ground the CT and see what happens.  Glad to hear the voltages look about right.

I'm slightly worried about the PT.  I'll do some more research and re check my wiring, but if anyone has any ideas why it is so hot please let me know!

 

[lassoharp]

I haven't done anything with the 6.3VAC CT at the moment - shall I ground this?

Yes.  I had a PS with filament CT disconnected for a while - sometimes it would hum sometimes it wouldn't.  It needs a permanent reference.

The Hammond 370x PT is also very hot.  I can touch it for a couple of seconds before it gets too hot to touch.  Is this normal?  It's the first time I've used one so I have no idea.

That sounds like trouble.  At that current draw that PT should be cool for a good bit of time.  Carefully recheck and confirm all your connections with the schematic.  With the voltages you gave it seems strange that the PT would be getting that hot - no evidence of excessive current draw.  Can you confirm to us how you have the 6X5 rectifier wired using the PT wire color codes and tube socket numbers?

 

[letterbeacon]

That sounds like trouble.  At that current draw that PT should be cool for a good bit of time.  Carefully recheck and confirm all your connections with the schematic.  With the voltages you gave it seems strange that the PT would be getting that hot - no evidence of excessive current draw.  Can you confirm to us how you have the 6X5 rectifier wired using the PT wire color codes and tube socket numbers?

Here's the Hammond wiring diagram for reference:

Ok I've just rechecked my wiring of the 6X5 and the PT and here's what I have:

Power transformer primary:
White + Blue wires connected to IEC inlet via a 1A fuse and switch
Black + Brown tied together
All other wires isolated

Power transformer secondary:
Green + Green 6.3VAC connected to V3 pins 2 & 7
Green/Yellow 6.3CT  not connected at the moment (but will be connected to ground when I get a chance)

Red + Red 240VAC connected to V3 3 & 5
Violet + Red/Yellow 240CT tied together and connected to ground

Yellow +  Yellow 5VAC isolated (not needed)
Yellow/ Black 5CT isolated (not needed)

Does that look about right to you?

 

[lassoharp]

Violet + Red/Yellow 240CT tied together and connected to ground

That may be the issue.  Only tie the CT (Red/Yel) to ground)  leave the 50V tap (vio) isolated.

  The 50V tap is probably a bias tap.

 

[letterbeacon]

Violet + Red/Yellow 240CT tied together and connected to ground

That may be the issue.  Only tie the CT (Red/Yel) to ground)  leave the 50V tap (vio) isolated.

  The 50V tap is probably a bias tap.

Ah yes, the thought crossed my mind as I was typing it out. Thank you, I'll try isolating it tomorrow when I next have the iron out.

 

[lassoharp]

Another way to look at it - you have 50V going directly to ground with only some of the PTs internal DCR in between - maybe several hundred ohms . . .

 

[letterbeacon]

So I've grounded the 6.3V CT and isolated the 50V bias tap on the secondary and... no more hot PT and no more buzz!

Kind of...

There is a buzz but it's so faint that I can only hear it when I turn the gain up full record a clip in my DAW and then increase the clip by 24dB, and turn the volume of my speakers way up, and only then I can hear it faintly in the silence between audio.  I'm not too worried by it, but I'm hoping to reduce it further by using shielded cable from the XLRs to the audio transformers (at the moment I'm using twisted cables).

Also now the 50V bias tap isn't going to ground, the PT is outputting a higher voltage so instead of getting 248VDC after R8 & R9 I'm getting 270VDC.  I've run out of resistors to drop the voltage to 248VDC for the time being so I'll have to look at that later.

The audio sounds like it hasn't got much top end at the moment which I put down to three things: 1) I'm using a Shinybox ribbon mic which are quite dark anyway 2) I may need to swap the polarity of the secondary on the input transformer as Doug recommended in an email and 3) I haven't put in the treble boost part of the circuit yet.

When I have a spare minute later this week I'm going to install RMAA and have a play with the polarity and treble boost circuit, but first I need to build some attenuators so the pre amp input sees the correct impedance from my Fireface.

Here are a couple of pictures of it so far - apologies for the bad camera phone, I'll take proper pictures when I've got a proper front panel to show off (this one's just temporary).

 

[emrr]

Nice build!

 

[letterbeacon]

Thank you, I'll get better ones up once I get the front panel done and have finished installing the JLM stuff.  In the meantime I've been messing around with measuring the frequency response of the pre amp.

Attached is a pdf with the graphs and findings of the pre amp after running some tests via RMAA.  As you can see the frequency response drops off pretty steadily after about 1k and before 100Hz.  Does this look normal?  I mean, as it is a tube pre amp I expected it to be a little 'dark'.

I tried adding the treble boost circuit to extend the top end a bit but didn't notice any difference.  How subtle is the treble boost supposed to be?

Another thing you'll notice is the 50Hz boost in the noise levels, which I assume is the mains hum.  I can try rotating the PT, but are there any other tips for reducing this hum?  I'm going to replace the twisted wires that connect the input XLR to the input transformer with shielded cable, but how much of a difference do you think this will make?

Thank you.

Hmmm, it won't let me upload the pdf -it says the upload folder is full.  I'll attach a picture of the frequency response and the noise levels.  The left channel has the BA-2C across it (via 75 ohm resistors across the + and - to bring the impedance up to 150 ohm) and the right channel is my Fireface looped into itself.