Altec 1567 Mods

[rickc]

I looked in my book of Altec schematics. I have an original 1567A data sheet, and C8 is 150pf.

That capacitor is unusual, being located in the bass control circuit. There is usually no capacitor bridging that resistor in the circuit... They must be correcting for something else that is going on.

Also, I've never seen a 1567 without the A suffix either in person or on paper. I've been watching Altec since the 60's.

 

[rickc]

c8 confirmed. I looked at an actual 1567 today and C8 is indeed a 150pf ceramic disc.

 

[Bill Wilson]

I was the sole technician for an Altec commercial sound contractor in early 1960's. I also started an on location recording co. for the first few jobs I rented a 1567 from the contractor. This was designed primarily as hi Q pa gear. The mike pres are a bit noisy when a  lot of gain is called for. They could be made quieter by removing the cathode bypass capacitor. This will cost some gain but will improve the sn ratio. you could convert those crummy tone controls to a  straight gain stage  to compensate for whatever loss is incurred by losing the catode bypass caps on the preamps.

Cathode bypass caps were the norm in early broadcast equipment in order to get more gain from a tube.
The increase in noise is not worth it as you have a class A input stage; will be much cleaner w/o.

As soon as I could I designed & built a 6 input mixer using UTC A 10 transformers feeding 1/2 mullard gols series 12AX7's that went thru a 100k log taper pot & thru a lever switch to select A,B or Center imaging; very clean stage. Was using 2ea. U47's, 2ea. sony C37's, 44BX & BK5A ribbons & AKG dynamics.

 

[lassoharp]

I was the sole technician for an Altec commercial sound contractor in early 1960's. I also started an on location recording co. for the first few jobs I rented a 1567 from the contractor. This was designed primarily as hi Q pa gear. The mike pres are a bit noisy when a  lot of gain is called for. They could be made quieter by removing the cathode bypass capacitor. This will cost some gain but will improve the sn ratio. you could convert those crummy tone controls to a  straight gain stage  to compensate for whatever loss is incurred by losing the catode bypass caps on the preamps.

Cathode bypass caps were the norm in early broadcast equipment in order to get more gain from a tube.
The increase in noise is not worth it as you have a class A input stage; will be much cleaner w/o.

As soon as I could I designed & built a 6 input mixer using UTC A 10 transformers feeding 1/2 mullard gols series 12AX7's that went thru a 100k log taper pot & thru a lever switch to select A,B or Center imaging; very clean stage. Was using 2ea. U47's, 2ea. sony C37's, 44BX & BK5A ribbons & AKG dynamics.

Hi Bill, Welcome to the forum.

Your tube mixer sounds interesting.  I just recently finished a 4 x 1 mono mixer with an aux channel and am interested at some point in attempting a stereo version.  Yours sounds similar to the Ampex MXs with the 3 position lever switch.

What machine did you use for tracking?

 

[gridcurrent]

They could be made quieter by removing the cathode bypass capacitor. This will cost some gain but will improve the sn ratio.

Cathode bypass caps were the norm in early broadcast equipment in order to get more gain from a tube.
The increase in noise is not worth it as you have a class A input stage; will be much cleaner w/o.

Respectfully, I question your assertion that removing a preamp cathode bypass capacitor will improve the sn ratio.  Perhaps improve distortion generation, but almost always degrade the noise figure.

One of the flaws that I see in the 1567a are the termination resistors across the input transformer primary windings.  Loading down the mic?  You won't ever see that in a classic preamplifier.

 

[Bill Wilson]

Hi Lassoharp: I used an ampex 351-2 studio recorder, it was mounted in a frame of square steel tubing that was supported inside a larger frame w wheels making it totally portable. The mixers preamp was based on a class A mike preamp design by Bill Dilley, owner of Spectrasonics. The channel selct switches were switchcraft bat handled lever switches, not the funky switches used on the ampex mixer. The L & R busses were fed from 100K pots thru a 100k series resistor. The busses fed into a gain stage followed by a cathode follower output; that fed the hi Z line inputs of the 351 tube electronics.

There were no vu meters on the mixer as I used the simpson meters on the Ampex. There was no EQ & microphone types were chosen for the instruments to be recorded. I did not record R & R or pop, so could get by without it.

 

[lassoharp]

Thanks Bill,

With the battery of mics you listed that sounds like a winning lineup for most anything. 

 

[tubologic]

They could be made quieter by removing the cathode bypass capacitor. This will cost some gain but will improve the sn ratio.

Cathode bypass caps were the norm in early broadcast equipment in order to get more gain from a tube.
The increase in noise is not worth it as you have a class A input stage; will be much cleaner w/o.

Respectfully, I question your assertion that removing a preamp cathode bypass capacitor will improve the sn ratio.  Perhaps improve distortion generation, but almost always degrade the noise figure.

Removing the cathode bypass cap will introduce cathode degenerative negative feedback in the AC (signal) path,this will result in lower stage gain,lower distorsion and lower noise. (The effect and purpose of NFB). Not a bad idea as the 1567 has way too much input gain and easily overloads with most modern mikes,it allways needed input pads anyway. Remember this was P.A (read:low-fi) gear with many design compromises,thus don't expect stelllar performance from a 1567,even with all the possible mod's. Better to save the trannies and build something from scratch. (I.M.O)

 

[gridcurrent]

"Removing the cathode bypass cap will introduce cathode degenerative negative feedback in the AC (signal) path,this will result in lower stage gain,lower distorsion and lower noise."

Please share with us your textbook references.

I set up an Altec 1567a,
test conditions:
150 ohms source Z, 600 ohms load Z, 117 VAC  60 Hz

Maximum Gain:  78.0 dB
Weighted (400 and 30 kHz filters) Noise:  -52.7 dBm
Equivalent Input Noise:  -130.7 dBm

With the input stage cathode bypass capacitor removed:
Maximum Gain:  71.6 dB
Weighted (400 and 30 kHz filters) Noise:  -58 dBm
Equivalent Input Noise:  -129.6 dBm

Lower noise?  Sure, but at the expense of gain.
The EIN will suffer if the cathode bypass capacitor is removed.

If you are questioning the maximum gain, it is because this particular 1567a has less gain partially due to the insertion of an unbypassed cathode resistor in the the 1st half of V3 (summing amp).

 

[tubologic]

Please share with us your textbook references.

RADIO DESIGNER's HANDBOOK, 4th Ed. Langford-Smith, p.309-310: Effect of feedback on hum or noise introduced by the amplifier.

Lower noise?  Sure, but at the expense of gain.

Of course, NFB will reduce the gain in the same amount and you need to increase the input level for the same output before measuring the S/N ratio. Obviously NFB has no effect on the thermal noise originating in the input circuit (microphone,transformer,input grid resistor...),this explain your measured figures which I don't contest. Nevertheless, NFB will help to reduce noise but the reduction will be less than the feedback ratio. The noise spectrum could also change depending on the bandpass/phase characteristics of the amplifier stage(s), making it more (or less) perceptible.

 

The EIN will suffer if the cathode bypass capacitor is removed.

If you are questioning the maximum gain, it is because this particular 1567a has less gain partially due to the insertion of an unbypassed cathode resistor in the the 1st half of V3 (summing amp).

I admit the gain/stage distribution in the 1567 is far from being ideal but I would not venture into a complete circuit re-design which is a labourious task.
Again,the ALTEC 1567 was adequate for P.A  but not really intended for studio work. If noise is an issue,selecting a good-quality input tube (12AX7) and replacing the resistors in the 1st stage with M.F types could help more than playing with cathode bypass caps.

 

[emrr]

I seem to recall many mentions of the change in stage loading when comparing cathode bypass with non-bypass, and the resulting effect on overall sonics. Possibly more significant than noise here.

 

[emrr]

Wow, too many 1567 threads.

Why is there a grid leak bias stage?  Haven’t sorted out if I’ve known why that makes sense here and forgotten.

 

[rackmonkey]

Why is there a grid leak bias stage?  Haven’t sorted out if I’ve known why that makes sense here and forgotten.

I’d never noticed that. Other than squeezing a little additional gain out of the stage, my assumption would be cost - trading what would have been an expensive bypass electrolytic back then for a smaller coupling cap.

 

[emrr]

I’d never noticed that. Other than squeezing a little additional gain out of the stage, my assumption would be cost - trading what would have been an expensive bypass electrolytic back then for a smaller coupling cap.

They are considered higher noise, so why in something that’s already a noise monster?  Gonna have to think about this awhile. 

 

[rackmonkey]

Yes, heater noise is often higher with grid leak stages than with cathode bias, for sure.

Had they not specifically advertised the unit as being suitable for recording/broadcast, I wouldn’t have thought the noise argument to be that compelling. But they did.

Interesting question.

 

[emrr]

One of the flaws that I see in the 1567a are the termination resistors across the input transformer primary windings.  Loading down the mic?  You won't ever see that in a classic preamplifier.

That's the other one I'd forgotten was happening here.  Altec liked to go with matched loading on preamp inputs far longer than anyone else.  It's usually a secondary resistance, primary here because of the multi-use nature of the input adapter socket.  I'm gonna bet most of those 180Ω loads are still in place, kind of hilarious.  Bye bye, I say. 

 

[emrr]

OK, RDH4 12.2(v) and associated reference 2.2(iii) seem to give some reasoning to use of grid leak with a high mu tube.