RCA Ba2A Pre amp qestions & answers

[emrr]

ba1a and ba11a are standard 40 db types. ba2 is 50 db. a few db fluctuation is normal to see.

ba1 rated for +10 dbm max output, ba11 for +18dbm.

 

[tardishead]

On a different note - I am going to build a BA-2c. The DC resistance for the primary of the output transformer is said to be about 2500 ohms.
If the transformer had a primary DC res of 470 ohms would this have an effect on gain and bandwidth in this circuit.
Is it essential to get this DCR in the right ballpark??

 

[emrr]

It generally suggests a lower AC impedance, and you would usually get something in a similar ballpark, say 1500-3500 ohms depending on particulars of manufacturers approach. 470 would certainly seem too low. The real answer is to put a 600 ohm load on the sec (assuming you know it to be designed for 600 ohms), take some measurements, do some math, and arrive at the primary impedance.

 

[tardishead]

Transformer is 20k to 600ohms.

But the primary is about 470 ohms DCR.

 

[emrr]

okay, now I'm on the same page. The RCA stuff tends to have a higher than normal DCR, when compared to other brands. Langevin is the same way. Not conversant enough to say specifically what's up, other than it would seem to be wound with much larger gauge wire, and have a simpler winding scheme. That is a larger than expected difference, so I'd say it's outside of the norm when compared to a similar UTC, etc. But no reason I can specifically raise for it to be problematic.

 

[tardishead]

Ok now onto the next stage. After getting a good frequency response from a RCA bA11A that I built I decided to use the transformers to build a BA-2c.

Preamp works. I am using 6SJ7 tubes again and 250vDC. And to my horror I found a bass rolloff that starts about 1khz -
(100hz is about -7db lower than 1khz.)

I checked all the components a few times and everything is up to spec.
After verifying that the input and output transformers were good enough response for the BA-11A I am now totally bemused. I biased the cathode of the second stage colder to see if that would effect it (pulling less current through the output transformer just in case it was being overloaded) but there is no change in frequency response.

Really confused. Do I need to load the transformers with resistors? There is something strange going on - I am wondering whether it is another difference due to the use of 6SJ7 instead of 6J7.

 

[emrr]

no, tubes don't have frequency response differences on that order. I had a BA-2A that had EF86's stuck in it with no circuit changes, and it worked fine. That's a fairly large difference compared to what you're talking about.

do you get flat response out of the transformers by themselves, out of circuit? I suspect the iron the most. Resistive loading won't help a curve like that, other than possibly being sure you have 680 ohms or so on the output side. That becomes more critical with low/no loop feedback circuits. But that still sounds like too great a roll-off.

It's possible the 27 dB of feedback in the BA-11A circuit was flattening out the response of the transformers, if you haven't tested them. The feedback curve may have happened to match a transformer curve.

 

[tardishead]

A few more observations.

When I turn the unit off and put a tone accross the primary of the output transformer I get an almost perfect frequency response through it.

I put a 600R resistor across the output and it improved a little with the mic pre turned on.

With the unit powered up I put a tone onto the grid of the second stage - I still get exactly the same roll off - so it has to be something to do with the second stage and output transformer. Why would the transformer work well on the BA-11A and not on the BA-2c????

 

[gary o]

Interestins stuff wish I understood all thats goin on....my ba1a ba11 ba2a all pre all worked fine with my one Sowter transformer, tho it is different to yours..I made a AWA pre with 6SJ7s & had fun & games so I would try 6J7 anyway bit of pain with different pin out tho, keep goin its gonna be right soon :grin:

 

[tardishead]

OK - I put another transformer on the output and it works as it should.
50hz to over 10KHZ is perfect 15khz is about 1db down. So I have 2 working preamps.

I was just trying to understand why the transformer would cope with the situation of the BA-11A but drop bass on the BA-2c circuit - the output stages are very similar. Maybe its something to do with the B+ wiring differences of my BA--11a and BA-2c.

On the BA11a I connected a regulated 250vdc supply to the first dropper resistor on the B+ line which I think from memory is 2k7. When it hits the plate it was about 230vdc.

On the BA-2c I connected the 250vdc supply straight to the output transformer (there is no dropper resistor after the power supply). When it hits the plate on this circuit it is 248v.

Is it possible that I am working at the limits of this transformer and too much B+ would compromise the performance. It works in my BA-11a circuit but I am just trying to understand the differences.

 

[emrr]

When testing with tone you should consider the source impedance. You are likely driving the output directly with a much lower source impedance than the tube provides (which improves transformer response), so you have to simulate the tube source with series resistance when testing if you want a reasonably accurate comparison. An accurate test of a 20K:600 transformer driven by a low-Z test source would involve 20K series resistance between source and primary, and a 600 load resistance on the secondary. I think that is the discrepancy you are seeing.

Transformer specs are not always as they seem. Consider many modern pieces that specify something like 10K:600, but then also specify a maximum recommended source impedance much lower than the impedance ratio. Many pieces require a lower Z drive, but do not tell you so, only assume you will do so. I've found 10K:10K iron from the 1980's that only approximated the stated frequency spec when driven with something around 100 ohms source Z. The assumption is for SS drive. OTOH, many vintage pieces specs mean exactly what they state, because the assumption is they would be driven by tubes having higher output Z.

The feedback loop in the BA-11A lowers the drive impedance of the output tube. The BA-2A has no such mechanism.

 

[tardishead]

Thanks Doug that explains it well.

Can you recommend any good books on tube theory that runs through basic amplifiers?

I just got a bucket of old Trimax transformers which sound awesome and I wanted to run them through their paces see what was good for which circuits.

 

[emrr]

the easiest to digest and follow IMO is the classic 1938 George Sterling - THE RADIO MANUAL 3rd ed. It does not get into feedback much though. See RDH for that, though much more intensive.

 

[craptical]

For anyones interest (in cap types etc), a BA-2C is on evilbay now, including picture of guts:

http://ainaz.pair.com/radioman/rcabc2ab.jpg

By the way: I´m thinking of giving this beast a try, using a Kenyon T-101 as output. Any comments?

/Dave

 

[emrr]

The Kenyon iron is very good. I don't know offhand if it takes DC or not.

 

[craptical]

Thanks Doug! :thumb:

The (very limited) specsheet says it´s suitable output for 56/76/6C5, 500 ohms sec. Doesn´t say anything about DC. However it´s listed in the 1941 catalog, was parafeed standard back then? If I understood you right in an earlier post 6C5 is almost a triode version of 6J7/6SJ7 (or at least interchangable in this circuit)?

Best regards

/Dave

 

[emrr]

DC or parafeed both possible all the way back to the 1920's. Safer assumption of course is if no DC listed, then not meant to have it. Probably have to try to know. 6C5 and triode 6J7 always on the same page of values in the tube manuals. I've heard highly opinionated claims that 6C5 sounds superior. I've never done direct A/B. Certainly easier to work with. Hum rejection and capacitance likely a bit different due to grid input point. Haven't looked at data sheets in awhile.

 

[CJ]

Kenyon made some really good stuff, and some not so good stuff.
Not so good, meaning lossy.

One model I tested drug down the sig gen so bad, I gave up.

UTC was sleeping with RCA for many years, thus, all the UTC iron on the BA stuff.
Some of it is private labeled, but it is still UTC.
'

 

[craptical]

Hmm... looks like I have to get my hands on a couple of 6C5s! :green:

No DC listed on any class A/AB/B output transformer in the catalog, only on modulation types (but they are rated at 150mA or more!). A schematic describing a transmitter with a mic input (in one of the old Kenyon papers) puts DC through an unrated (interstage) transformer, though. Well I´ll build a prototype and check it out. If the Kenyon behaves badly on DC, are there any direct problems with trying parafeed (except the obvious; a large cap in the signal path...)? Which UTC is the "correct" iron for this circuit?

Thanks!

/Dave

 

[CJ]

Here is an old RCA, does ity look similar to yours?
If so, maybe the iron would sound great.
LS 27 is the big boy.
LS 10 is kind of pricey.
Same exact xfmr as the A-10, only wound on a bigger core.
And more shielding.