How amplifier switching input jacks and “interactive” volume controls work

switchingjacks
Switching Input Jacks schematic

There are some very simple circuits found in guitar amplifiers that we take for granted. But at the same time, do not understand. And there are other things thought to be “magic” , again, because they are not understood.  (You can click on these images in order to see them better.)

This is an EASY article,  brought on by my recent purchase of a 1967 Blackface Bassman amp that was modded to the first (and best) circuit, the AA864. –- Keep reading.

Two input jacks – found on Fenders and Marshalls (The Marshalls in question have four jacks, the Fenders usually have two, but the Bassman amps have two in each channel.)

Why is one high gain and the other low? And how does the whole thing work, what are these resistors doing?

Overall the two high and low gain jacks are a very good trick using switching jacks and as few parts as possible.

SwitchcraftMonoTipShuntL12A
Mono Switching Jack

Jack #1, the high gain jack has a 1 meg (1,000,000 ohms) resistor wired across the jack from the hot tab to the ground tab, and the grounded lead of the resistors is bent back and soldered to the middle tab – the “switch.” The switch is grounded on jack #1.

These switching jacks have an extra “leaf” that is touching the hot, or “tip” connection of the jack – but only when nothing is plugged in – called a “normally closed” jack. This can be made to do all kinds of things.

SwitchcraftMonoTipShuntL12ADiagram
Mono Switching Jack Diagram

Trick #1. In the case of the #1 jack, when you do not have your guitar plugged into the amp (either jack) the “switch” is closed – since it is grounded, the amp is turned off for low noise, and no chance of runaway.

The 1 meg resistor is the “grid load” resistor for the first tube stage of the amplifier. All tubes have to have a grid load resistance or they will not operate.

Instantly many of you are thinking, “what about the guitar pot, or the pickups, aren’t they a grid load?” Yes, they would be – but – if the guitar had active electronics, or had a capacitor in series with the output signal from the pickups, there may be no direct resistance from the guitar. It is better to be safe and have the resistor there.

Ok, more on the input jacks. The second jack has it’s switch soldered to the first jacks hot lead. Each jack has a 68k (usually) resistor from its hot lead, and they join together to make one lead to the tube stage.

These “series resistors” are sort of input buffers – to keep the amp from getting high level spikes from the instruments. Not totally necessary for the vintage guitars these amps were made for, but a real good idea for modern guitars with high output pickups, and in the old days, accordions!

Here is trick #2. Jack #2 has it’s switch soldered to Jack #1’s hot connection, and a 68k resistor on it’s hot lead – when the switch is closed – nothing plugged into jack #2, it’s 68k resistor is placed directly in parallel with Jack #1’s 68k resistor, for half the value, 34k. Less series resistance.

Trick #3. If you plug into jack #2 only, check out what happens. We already learned that if nothing is plugged into jack #1 the hot is shorted to ground via the switch – so now Jack #1 represents a direct ground connection.

BassmanFrontPanel
65 Fender Bassman Front Panel

The signal from jack #2 would pass through the 68k resistor, and at that point see the 68k resistor from Jack #1 with its other end grounded!! So the signal has 68k in series (more resistance, a lower signal level gets to the amp) and a 68k grid load resistor – much lower resistance than the 1 meg. A lower value of resistance to ground reduces the signal to the tube (again.) Check out the schematics to see what is going on. (Click on it to see it full-size)

bassman_ab165_schem
Fender Bassman Schematic

The mysterious “interactive volume controls” on the Tweed Fenders (Bassman, Twin, Super, Bandmaster, Deluxe etc.) This is not some magic trick, or even an intentional feature – it is a minor design error.

These amps have two volume controls each controlling one stage of a 12AX7 (7025) tube. They are very simple and economical circuits, only 5 parts per circuit. Each output comes from the volume control stage and passes through an “isolation” resistor – a resistor placed in the circuit to isolate the two volumes from each other, and at the same time to mix them for the next single input tube stage.

The “problem” or “feature” if you wish to call it that, is that the isolation resistors values are too low (270K on Tweed Bassman amps, and even lower – 220K on Blackface Bassman amps.) They do not isolate the stages from each other enough – one stage can still “see” the grid load resistance from the other. It appears to the channel you are plugged into, that there is another resistor in the circuit formed by the other unused volume control and it’s isolation resistor.

The real “magic” of it all is probably one of the absolute best “lessons” you can learn about vacuum tubes. You would guess that the best and loudest tone would be with the highest resistance, with the unused pot turned all the way up, but that isn’t so. It is somewhere near 5 or 6 (on old amps, a new amp may have different pot tapers and it could appear at different points.)

When you are turning the unused volume control you are “fine tuning” the grid load resistor to the next stage of the amp. An optimum (best sounding) value is not all the way up, but somewhere near the middle – the lesson? Tubes are NOT linear or mathematical devices – they do not respond to exact mathematical figures and have “hills and valleys” in their response – maximum performance can not be calculated directly – you have to hear it and you have to be aware that the “truth” may not be at some exact calculated place, but at some far more imprecise figure.

New server IP address

Due to the migration to Frontier from ATT, the server’s static IP address has changed from 108.89.65.174.  It’s new IP address is:  32.220.54.46.

Also, the web pages at cap.vapor.com have been migrated to cap.crabradio.net, so that they fall fully under the crabradio.net domain, owned by me.

The subdomain webpage still has a cap.vapor.com graphic on it, but I’m hoping my friend The Moonshiner can come up with a similar graphic to just drop in.  😉

The only problems that you will encounter is if you attempt to load cap.vapor.com, as that dns is dead and unlikely to be updated.

I’m still waiting for the rest of the arcnet irc admins to update the US1 server IP address, and I shall reconnect us1.arcnet-irc.org to the rest of the servers.

Thanks for your patience.

Server sendmail and icecast info

Well, I just noticed that the podcast and regular blog weren’t sending out email notifications any longer.  Pretty easy to track down.  For some reason, sendmail was not running.  Even easier to fix.  I simply did (as root):

# systemctl start sendmail

Tested the podcast’s php mail function, and all is well.

Something else that I’ve been chewing on for awhile and finally whipped a couple weeks ago.  I wanted my icecast server and IceS streaming source to behave like centovacast, with a ‘live’ mountpoint that ‘takes control’ of the stream…..  All I had to do is add these lines to /etc/icecast.xml:

<mount>
<mount-name>/stream</mount-name>
<fallback-mount>/live</fallback-mount>
<fallback-override>1</fallback-override>
<hidden>1</hidden>
<public>0</public>
</mount>

<mount>
<mount-name>/live</mount-name>
<fallback-mount>/autostream</fallback-mount>
<fallback-override>1</fallback-override>
<hidden>1</hidden>
<public>0</public>
</mount>

<mount>
<mount-name>/autostream</mount-name>
<fallback-override>1</fallback-override>
<hidden>1</hidden>
<public>0</public>
</mount>

Then, I just had to set IceS to send to the autostream mountpoint, instead of stream.  So, now, in absence of live, when you connect to the normal stream mountpoint, you fallback to live, then fallback to autostream.  The fallback-override directive allows a live source to ‘pull back’ listeners connected to the autostream.  😉

 

Security cameras now running under Fedora Linux (64 bit)

Well, I managed over the past few days to get the security cameras up and running on Fedora 21 Linux 64 bit using Zoneminder.  The cameras were running under WebCamXP under Windows 7 64 bit, but I grew tired of Bill Gates deciding to reboot the server, even though I instructed it not to perform updates.  Microsoft would occasionally decide that it was in my best interest to apply an update automatically, and do a reboot, rather than allowing me to do that.

Click here for some screenshots from the zoneminder site of it in action.  I’m already a fan of it since it’s running under Linux and Zoneminder Montageutilizing mysql.  For right now, it’s running ‘in parallel’ to the Win 7 server, to make sure that there aren’t any issues.  In a week or two, I shall finalize the migration.  Another point that would have probably been an issue with running two security servers ‘in parallel’ would have been overloading my wireless network, so I converted everything to ‘wired’ with the two servers in a temporary location to make sure everything is working properly.  I left the wireless access configured for it to use it as a ‘fallback’ in case the wired ethernet fails.  (The picture above is not my actual cameras, btw.)

Another upside to this is that the zoneminder console is far less cpu hungry than its windows counterpart.  It’s running at around 75% to Zoneminder Console85% cpu free.  For you hardware geeks out there, it’s running Fedora 21 (64 bit) on a compaq (or HP, if you would rather) machine with an Intel i5 quad core 3.10 Ghz cpu with 8 GB of ram.  (I checked, and got similar numbers on Fedora 21 32 bit with 4GB of ram.) (The picture above is not my actual console, or web address.)  (Yes, I have my console password protected.)  If anybody is inclined, and needs help to set things up, just ask, and I’ll be happy to post a tutorial here.  (Since it’s linux-based, its hardware requirements aren’t that high.)

Tech, music, and general ramblings…..