1930's microphone preamp

[Conviction]

So here's an odd one: a microphone preamp from the late 1920's/early 30's. I actually found three of them in a storage, all with slight circuit variations (most likely due to different tubes used). I haven't had the time to confirm if its just a matter of biasing.

One of them actually works - and sounds unexpectedly good. The preamp's designed to be used either with a so-called Reisz microphone or a dynamic.

The AGA (manufacturer of some great broadcast & film equipment) input transformer looks like a rare one. Some of you have  probably encountered Ferranti transformers by now, but they're getting pretty scarce too.

Here's the schematic:

Does anyone recognize the circuit? Any similarities with other preamps from the same era?

The one working sports 2 Telefunken RE124's instead of B405's and, or you can see in the schematic, B415K's. Should be fully equivalent. They all have a B405 in the output stage.

 

[PRR]

> Any similarities with other preamps from the same era?

If this is late 1920s, there was not a lot of choice of circuit.

You get a gain of 10, at most, from each tube. You didn't want to close-talk these mikes. So you start off with three tubes.

The last tube is logically a small loud-speaker tube. B405 can output maybe 250mW (+28dBm), though much less at low THD (you can have 10% THD one place in a system, but not at every amp in a long broadcast chain).

No unipotential cathodes yet, so self-bias was awkward. So fixed-bias, tapped from the filament supply.

Filament supply has to be clean DC (and was surely a lead battery).

The 4V filaments on 6V battery suggest that for these tubes in home use you had a filament rheostat. Turn down when battery is fresh, cut out when battery is drained, to hold about a 4V glow in the filaments. For broadcast use the 6V was maintained pretty close (either generator online or re-charge every night) so the 6V-4V droppers are fixed. (And also used for grid bias.)

The gain control is a switch because resistances were still new and high-resistance "wiper" potentiometers were very noisy. Stud-switches were mature and reliable (if polished often).

One stage of gain before the gain-pot is still a popular design. Pot-in-front means excess hiss in most situations; two stages of gain risks overload before the pot can turn-down.

Also you do not put loss between driver and output tube. The gain/power compromise means the final tube needs large drive signal, often about all the driver can make.

The small oddity is the 2nd transformer. Perhaps B405 had a very low maximum grid resistance. Maybe in days when high-value resistors were new, the cost or reliability of mature transformer technology looked better than it does today. (It is not likely the B405 is driven Class A2, grid current; though perhaps an overload sounds better with the plate:grid transformer.)

The AF5C is labeled for DC in the primary but is AC coupled here, probably for better bass.

B405 is self-biased with the 400r resistor. (Actually the two B415s also flow in the 400r, but they are a small fraction of the total.)

If the OPM/c OT says "2.7/1", and we assume 600r load,the ~~4,400 primary resistance is near-perfect for a B405's ~~2K plate impedance.

Excellent midrange was common by 1929 (in professional audio). Long-long chains of transformers (and offset speakers) showed problems with tap-dance, while being "perfect" on speech and the music of the day. Of course home radio transformers were low-price and low-fidelity (and for many decades).

 

[emrr]

More pics?

A claim of 3.5:1 on the AF5C from an ebay auction.  That does look odd if that is the orientation, agree with PRR on grid value. 

I'm not used to seeing bias from filament battery tap rather than a dedicated C battery.  Is this a portable unit? 

I have wondered about the specifics of Ferranti US versus Europe.  Was there separate dedicated US manufacture and product lines, or were they distinct?  Not a lot of info out there. 

 

[PRR]

> More pics?

+1!!

> filament battery tap rather than a dedicated C battery.  Is this a portable unit? 

Perhaps; but in any case, filament tubes for small audio "must" run on battery. Otherwise the hum is impossible.

(A well-known exception is 2A3; but it is a BIG tube with BIG grid swings. And only 2.5V, and center-tapped when fed AC. Assuming 10% of the 1.25V each side gets into the audio, S/H is 33dB; cancellation is probably better than that; nevertheless hum is an issue for 2A3 in modern expectations.)

Also broadcast thinking came from telephone systems. Battery is dead-clean and works through short power failures.

I agree that tapping A to get C is semi-novel to me. US designers often threw in a C batt. But here a happy lead-voltage leaves a couple volts to spare, and V1 V2 only need ~~1V of bias. And eliminating a third battery (and connections and checking and replacements) is a worthy goal.

I too am curious about AF5C ratio, also the apparent extra resistor on the gain switch. Almost like 2 tubes was not enough, 3 tubes was too much, and they threw-away gain to hit the max-gain they wanted.

Recall that Harold Stephen Black developed his negative feedback 1927-1934. While today we think it is the obvious way to control amplifier gain, it was not obvious at the time (not even well accepted). Considering the era of Conviction's preamps, NFB was not yet in the toolbox. So adding loss in pad and transformer may have been the "obvious" way to get 2.4 tubes-worth of gain.

Also a ~~3:1 driver transformer semi-matches the overload points of V2 and V3. This can cause a cancellation of even-order distortion. Perhaps from 5%-10% for V3 alone to 2%-4% for V2+V3. The odd-order distortion rises but not objectionably (for the era). The Old Guys were clever in ways we have forgotten.

 

[emrr]

a ~~3:1 driver transformer semi-matches the overload points of V2 and V3. This can cause a cancellation of even-order distortion. Perhaps from 5%-10% for V3 alone to 2%-4% for V2+V3. The odd-order distortion rises but not objectionably (for the era). The Old Guys were clever in ways we have forgotten.

I hadn't thought of that.  Interesting.  I suppose this would relate in preamps 10-20 years later which still skip loop NFB, many seem to have close/identical headroom in both of two stages.

 

[PRR]

> preamps 10-20 years later which still skip loop NFB

Some of those also used relatively more-beefy output bottles, so didn't have to run so close to overload, thus got acceptable THD without NFB.

But yes, trying to null final against driver THD is an old trick. Especially useful with low-gain power tubes.

 

[emrr]

Experience behind the wheel suggests that Collins Radio used this trick a good bit before 1950. 

 

[Conviction]

Here goes. Sorry for the crappy quality, took it in a rush.

The 4-pin female connector indicate that this unit was used in early line audio.
I don't remember the name of the jack, but it was the telephone standard till (IIRC) the late 1940's.

Interstage & output transformer:

The input transformer:

Close-up #1 of the interstage.  Unfortunately, this is all that's printed on it:

Close-up #2 of the interstage.

I actually found a couple of spares if someone would like to practice a little dissection. I wouldn't mind.

 

[Conviction]

I just realized the pics show different units... must've mixed it up. But they're almost identical, except the gain control. Two of them have the 5-step, the other 12. I'll post the latter version in a bit.

 

[PRR]

> the jack, but it was the telephone standard till

In most areas, "pro" audio adopted as much telephone technology as they could use.

The connectors were good-quality so they could survive hundreds of re-connections a day. Telcos bought them by the many-thousands, so price was not too high. Western Electric would fulfill orders from anybody, and in most other places the connector company was probably happy to increase sales.

In radio they were always interconnecting to telco lines for remote broadcast. Football game won't fit in the studio. Mayor opens the new library. Pope comes to town. In England, radio and telephone reported to the same top-boss. In the US, radio operations were separate from telephone but they had to work together all the time.

 

[ruffrecords]

In England, radio and telephone reported to the same top-boss. In the US, radio operations were separate from telephone but they had to work together all the time.

In the UK, radio was the responsibility of the BBC. Telephony was the responsibility of the GPO (General Post Office) so post and telephony were under one head but radio was quite separate.  The GPO had a monopoly on copper connections so the BBC had to use them for all interconnects between studios. When the BBC later opened local radio stations, cost constraints forced the BBC to look for alternative solutions for some functions. For example, most BBC local radio stations rebroadcast the national radio channels. I remember visiting BBC Radio Nottingham in the 70s and seeing the bank of Leak Troughline FM receivers they used to  pick up the national channels for rebroadcast.

Cheers

Ian

 

[emrr]

On #4 the Ferranti appears to have been added later. 

 

[Conviction]

On #4 the Ferranti appears to have been added later.

Good catch. Perhaps they found the Ferranti to be more suitable very early on. I wonder what it replaced.

Also, looking at the schematic it seems that the SS1794-5 OT (unknown manufacturer) got replaced by the Ferranti OPM/C somewhere along the line (#50). Perhaps the SS1794-5 and OPM/C are equivalent.

The Ferranti OPM looks much older though, which makes me a little suspicious.

 

[PRR]

> radio was the responsibility of the BBC. Telephony was the responsibility of the GPO (General Post Office)
> ... post and telephony were under one head but radio was quite separate.

GPO was (is) the "licencing authority". After initial chaos (following the same path to chaos as US radio), GPO pulled-back and issued a single license to a company to be known as the BBC.

This could have gone very wrong. By combination of brilliance and luck, it served very well, at least til long past the era of the preamps in this thread.

Yes, quite correct, this made an arms-length relation between GPO and BBC. Both could deny blame for the other's actions. But clearly they "had" to co-operate on every little thing, from an island-wide audio network to the plugs used to connect. BBC didn't have to use GPO plugs, but they would need a lot of adapters otherwise. GPO could not refuse the BBC access to their plug supplies without good reason. Government clearly wanted an orderly no-problem radio system. If snitty attitudes caused a Minister, Parliament, or King to come down and de-poop the fan, which-side of the License they were on wouldn't save them. (If BBC wouldn't fire John, then de-license the BBC and de-authorize their set-fee funding.)

The UK GPO also operated the telephone system. In the US, the government authority did not "operate" anything. FCC blows with the political wind (tho often out of sync). The telephone companies were semi-chaos locally, but Bell System was far ahead on Long Distance. By the 1930s, Bell made more than half their money on national radio networks, not people calling people, so they got very cooperative. OTOH, local telcos could be slow and surly (and expensive) about setting up lines for radio, because their local regulators only looked at "people" service quality/cost, not the needs of industry or radio.

You also find extensive use of telephone technology in early Hollywood movie audio (weakens my top-boss argument). In part because Ma Bell saw movie-sound as an extension of their main work, and because they were far ahead of all other sound parts producers. Bell failed to produce a really-good theater speaker (even though they had some) so the studios developed their own. Anti-monopoly sentiment got the government to tell Bell to stick to telephones and break-off their movie-sound business (so RCA and others could try for a share), but the equipment didn't change and some of it still made by WE (not quite captive of Bell).

> bank of Leak Troughline FM receivers they used to  pick up the national channels for rebroadcast.

Hmm. Is that a later thing? I was under the impression that in the 1930s, much audio was carried over radio-grade telephone lines. Yes, what works for early AM when all of Europe is cross-modulating the home set would not do as radio quality improved later. We used a Fisher 500 to catch Allison Steele off an FM station to cover gaps in the schedule. (We had a note said it was OK.) We could *see* NYC from the top of the tower so reception was excellent. And my arm still hurts from nailing that FM antenna into good concrete.

 

[ruffrecords]

> bank of Leak Troughline FM receivers they used to  pick up the national channels for rebroadcast.

Hmm. Is that a later thing? I was under the impression that in the 1930s, much audio was carried over radio-grade telephone lines. Yes, what works for early AM when all of Europe is cross-modulating the home set would not do as radio quality improved later. We used a Fisher 500 to catch Allison Steele off an FM station to cover gaps in the schedule. (We had a note said it was OK.) We could *see* NYC from the top of the tower so reception was excellent. And my arm still hurts from nailing that FM antenna into good concrete.

Yes, much later, in the 70s. You have to remember that commercial radio did not begin in the UK until the 70s. Before that, the only local radio stations were operated by the BBC. Back in the 30s it was the four national BBC stations and that was it. There was little if any local programming so the need for audio lines from sports events , for example, was rarely required, the exact opposite of the States.

Cheer

Ian

 

[Conviction]

Here in Sweden, Televerket/Kungl. Telegrafverket (Royal civil service) owned the telephone lines up till the early 1980's. Sveriges Radio/SR (the Swedish equivalent of the BBC) worked alongside as best they could, but were often limited due to all the bureaucracy involved. They did however use the telephone lines a lot.

(Remember these? http://groupdiy.com/index.php?topic=54451.0 During the 50's these, I found out recently, were the standard amplifiers for returning signals to the station).

Looking at the schematic below, which I found in the same storage, I'm certain that these preamps were infact made and used by Televerket. This microphone preamp from 1945 is called "Type 43". Creative enough.

 

[PRR]

> I'm not used to seeing bias from filament battery tap rather than a dedicated C battery.

Example from a different class of unit:
http://groupdiy.com/index.php?topic=54382.msg768707#msg768707

Schematic shows grid bias taken from heater supply. (Condenser mike-- battery or added wire is awkward.)

Does assume that heater/filament voltage tracks with plate supply, more or less. In professionally maintained gear, this is reasonable. In beach-radios or small-market remote boxes you could some day have a fresh B batt and a stale A batt, high B current and poor output.

 

[DaveP]

I didn't realise that Swedish was so similar to German:- Forstarkare, =verstarker =amplifier.

A very nice schematic, thanks for that Conviction.

Best
DaveP

 

[PRR]

> Swedish was so similar to German:- Forstarkare, =verstarker =amplifier.

20th-century words are more likely to be adopted from a similar language, by reading books and magazines.

Old-old words are mostly more different. Even if the Swedes got the word for "cow" from Germanic people, they probably say/spell it very differently now.

 

[Sredna]

Not really.

German for cow: Kuh
Swedish: Ko

Pronunciation is almost identical.  [kuː]

Swedish is also a germanic language and before the end of WWII German was the first foreign language teached in school, now it's English.  8)

The last part of the word Förstärkare is "Stärkare" (Germ. Stärker) and comes from the adjective "Stark" which means strong.
Förstärkare = a thing that makes (audio) stronger.

Sorry for OT.