Teac 2A mods

Recording Techniques, People Skills, Gear, Recording Spaces, Computers, and DIY

Moderators: drumsound, tomb

Post Reply
User avatar
buyin' a studio
Posts: 895
Joined: Tue Jan 03, 2006 1:39 pm
Location: in exile

Teac 2A mods

Post by floid » Fri Feb 15, 2019 4:27 pm

The latest in my series entitled, "let's put some polish on that turd!"
This is a funky little barebones mixer meant for use with the 3440. Six channels, four busses, and not much else. But it looks cool, and... well, it looks cool. That's enough.

Reworking the channel
Each channel has an unbalanced 1/4" mic input. I balanced the first five using transformers i scavengeded from a TOA p.a. mixer. Channel six was left stock for use as a DI.
I opened the chassis holes to accomodate xlr panel jacks using a step bit, doing the last two steps by hand and ratchet to avoid tearing the chassis - the spacing is pretty tight. I had to turn the panel jacks upside down so the release catches were't in the way of the line inputs. The two center wires on the input header connected to the t.s. mic jack, these were twisted, shrinkwrapped, and soldered to pins2&3 of each xlr. I laced a ground bus and tied it to chassis ground via one of the xlr panel bolts.

The channel cards have a ground plane directly behind the bus switches with enough room to drill holes for transformer mounts and wire pulls. I spliced on lengths of shielded cable to the transformer leads, cut the traces to pins 2&3 on the header socket, and tacked the primary leads to these points, the secondary leads to the cut trace on R1 and ground.

The transformer added roughly 14dB of gain to the mic input, and 10dB to the attenuated input. After playing around with things for a while, i decided this was too much: on drums, even with the preamp attenuated, the faders had to be pulled down quite a bit with most mics, and things sounded dull, as if the pre were bogging down. Frequency sweeps showed a loss of high end at high gain. I trimmed the values of the main feedback resistor R5, and attenuation resistor R7, which is switched in parallel. Experimentation led to dropping R5 from 100k to 47k, and R7 from 10k to 5.1k. This yielded 4dB additional gain over stock and maintains a near perfect 20 dB step attenuation, while working the preamp less. I tried a few other transformers and found these values would be dependent upon type.
The feedback loop ties to ground via C3 and R4 in series. This might be an alternative point to experiment with gain structure - see the 312. I reduced the value of C3 from 47uF to 10 uF to bring the hi-pass action at this point from roughly 3Hz to roughly 15Hz, as it seemed to improve the tone a touch.
C1, the electrolytic mic input cap, was traded for a 4.7uF wima - a smaller value than the stock electrolytic, but still plenty large to keep the filter well below audio frequencies. The line in cap was traded for a 1.5 uF wima to similar effect. All other electrolytics were replaced with 47uF/25v types. The green chiclet films in the bass section were replaced with wimas, the treble will have to wait until i get around to getting the proper value.

I also applied my patented super secret eq mod - with the eq knobs removed, i ran frequency sweeps and tuned for flattest response before replacing the knobs with this as "center". All channels were the same: bass one notch up, treble one notch down.

Reworking the bus
My drawer of funky iron also had some Tamura GA82830 line transformers from one of the lesser PM series Yamaha mixers. I'd put iron on the ins, now i wanted some on the outs.
The block diagram for the 2A looks something like this:
mic/line channels > mix bus amps > insert > four gang "master" bus fader > bus line amps
The send point and final output are the same level. The only "advantage" to using the line amp is having the master fader - which, frankly, would be a lot more useful if there were four discrete faders - and this comes at the expense of fader loss and an extra gain stage. I saw neither the real estate nor a pretty way to turn one fader into four. So i decided to place transformers on the sends and use these as the primary outputs, with the option of bridging the inserts on busses that needed some padding after driving the iron.
To do this i mounted a piece of scrap angle aluminum at the front of the mixer and mounted the transformers so they stood between the channel cards. I replaced the leads from the mix amp card with shielded wire, and brought down more shielded wire from the send jacks. i laced these to a piece of perf board, and pulled the transformer leads thru as i brought it down onto standoffs mounted to the transformer bolts. This part was a pain in the ass, but i couldn't think of a prettier way to terminate everything without having problems with strain relief as i moved things around. Those transformer leads are dinky.

Reworking the power supply
The power supply is a cheesy little thing delivering +/- 13v. Or actually, more like a generous +/- 12v. It does this via a 34v power transformer rectified into a 7824 24V regulator with pin2 tied thru the power indicator LED to ground. The LED floats the regulator a diode drop above ground, bumping the output voltage to 25.1v. A series "matched" pair of caps and resistors creates a "center" tap that is used as the mixer's ground, while what the rectifier and regulator/LED think of as "ground and B+" are used as the +/- supplies.
Which works, sort of. The primary rectifier cap C1 is rated to 50v, barely above the 48v peaks the transformer produces. The regulator is mounted directly to the pcb, where it drops a lot of voltage in the form of heat. Low tolerance components form a "center tap" that is only approximate, and this method will never deliver a TRUE center. And the rest of the board would happily run at higher voltages, if only the stock caps weren't rated to a measly 10v.
This isn't the power supply this mixer needs. But it's probably the one it deserves.
C1 gets bumped to 63v. The 7824 regulator is physically raised from the board and mounted to a heatsink with thermal compound. The trace from pin2 to the power LED is cut to add two blue LEDs in series, raising the regulator further above ground and reducing its voltage drop/heat. The center tap resistors are replaced with precision types. The mixer now operates at +/- 15.2v with a "truer" center tap and bonus super secret hidden under the hood bling.

Putting a skin on it
Not much to say here. I had some scraps of walnut and poplar. Danish oil finish. If i had it to do again I'd have put a coat or two of clear on the poplar before assembling the top panel so it didn't darken down as much.

So there it is. Keep an eye out for my next installment: "no phantom, no polarity, no midband: why does my record sound like walnuts?"
Village Idiot.

audio school
Posts: 8
Joined: Mon May 08, 2006 1:58 pm

Re: Teac 2A mods

Post by mds » Fri Feb 15, 2019 11:32 pm


User avatar
zen recordist
Posts: 8900
Joined: Tue May 13, 2003 8:08 am
Location: Chi

Re: Teac 2A mods

Post by vvv » Sat Feb 16, 2019 7:21 am

bandcamp; vlayman;
THD; Geronimo Cowboys;
I mix with olive juice.

User avatar
re-cappin' neve
Posts: 724
Joined: Mon Apr 25, 2011 11:33 am
Location: Cleveland, OH

Re: Teac 2A mods

Post by markjazzbassist » Mon Feb 18, 2019 7:00 am

i love it, how does it sound? what are you going to use it for?

User avatar
carpal tunnel
Posts: 1591
Joined: Tue Jun 01, 2004 5:47 pm
Location: sf bay area

Re: Teac 2A mods

Post by shedshrine » Mon Mar 11, 2019 5:32 pm

That must feel so good to be able to bend circuits to your whim like that.
Rock on good sir.

Post Reply

Who is online

Users browsing this forum: No registered users and 44 guests