Noisy Teletronix LA-3

[squib]

so the mystery remains on why there is DC on that second winding on the transformer! Any other ideas??
This unit is very good, after all it is a LA2a with a discrete A/B amplifier rather than a tube amp. The one i modified as per my earlier post gets daily use in a studio. I also moved the comp / limit switch to the front panel and tweaked the value of R3 to get the best limit ratio. Raising the input impedance by removing the pad at the front end also makes it play better with the other children!

 

[Gus]

R.G Keen www.geofex.com has done research into the early vox whas. What do whas have to do with this? The early Vox whas can sound very good and the sound seemed to follow the inductor. He measured the inductor and found it had become magitized a little, this is sosmething you don't want and should not happen. It is thought the cheap materal used for the core was a Happy accident.

IIRC the cheap core after years of use generated a little more 2nd harmonic. R.G.'s writeup was the first I read of something like this.

I wonder what happens to the sound with the winding in or out of LA-3 circuit. How does the harmonic spectrum change.

this is the best I could think of so far.

 

[Basspumper]

Yo Guys,

I tried the Ed Evans Mod and finally the noise is gone !
A nice clean sound without any hiss : I've been waiting for that for the last 5 years.

Many thanx Roggy and SPG !!

 

[jhaible]

[quote author="squib"]so the mystery remains on why there is DC on that second winding on the transformer! Any other ideas??
[/quote]

To compensate for the DC current that comes from the amplifier.

At first, the transformer looks capacitor coupled, but this is only true for the output stage. The _driver_ stage (Q4) sends some DC current into the transformer via R16. This is compensated by a DC current thru the second primary.

Quick calculation: The base of Q6 is approximately at 1/2 Vb. This causes a DC current of 0.5 Vb / 1.2k (R16). This is compensated by the other path with a current of Vb / 2.4k (R20) thru the other winding.

JH.

 

[mesmer]

Heres the evens mod:
LA3A Mods The original LA 3A (and LA 2A) were designed as broadcast compressors. The gain structure and facilities of the unit were arranged for that purpose. These units have found wide spread use in the recording studio, and as such require modification to better be utilized in the studio environment.. 1) Relocate Compress/Limit switch to front panel For convenience. 2) Add Link Switch to front panel. For convenience. 3) Add HF Boost switch to front panel This facility was originally used for broadcast transmission systems utilizing a pre-emphasis audio curve. This facility was designed to keep these EQed transitions from overloading in the HF region. Though not designed as such, in recording studios this function is used similar to a deesser, making the compression circuits more sensitive to HF (esses) signals. 4) Set HF boost rear panel control to MAX. FULLY CCW. 5) Relocate Power Switch to front panel. 6) Add BYPASS switch to front panel. Install termination at BYPASS switch (see below) 7) Align LINK facility. 8) Set rear panel gain switch to "50". This removes the PAD in front of the input transformer. Making better signal to noise ratio. 9) Add 15k resistor in parallel with R14 (220k). equivalent resistance = 13.5k This reduces the gain taken in the front end by leaving the gain switch in the "50" position. Both these together also put the GAIN and PEAK REDUCTION controls in a better range for interface in the recording studio. REMEMBER... Even if you don?t do all the other mods, Terminate the output transformer with 600 ohms (620 is ok) to prevent frequency response and transient ringing anomalies .. This of course assumes that the circuit that the LA3A will be feeding is NON terminating... i.e High Input Impedance (most circuits after 1970 are). Ed Evans Circa 1991, update 2003

 

[Kev]

[quote author="mesmer"]LA3A Mods The original LA 3A (and LA 2A) were designed as broadcast compressors. [/quote]

LA .. Levelling Amplifier ( leveling ? )
and yes, designed for broadcast.

good post there mesmer but I think it could have been better.
very presumptuous of me I know

are you interested in having the bulk of that appear on the DIY Factory pages as either in the data section or as part of an LA html page ???

let me know ... I think it is worthwhile ... anyone else have something we can cross link with this info ?

 

[Guest]

Kev-

That looks like the text from the pdf file I have here(minus the schematic excerpt demonstrating the mod). If you want, pm me your email and I can send it over to you.

Zach

 

[Kev]

yes

please do
this could be the PDF that circulate a while back over at PSW ... with John K. or was it with .. :wink: .. grumble bum ?

If this thing has a contact name perhaps I can post it 'complete' at my new place.

 

[zmix]

Here is my take on the various LA-3A mods.

On my LA-3A, original in every respect, and serial number 4093 (anyone know the date on that?), I have found that this thing will put out an incredible 40VAC p-p into a 600 ohm load at clipping. That is a lot of signal in a digital "18dB of headroom" world.
At the 50dB (high gain) setting and the gain knob turned fully clockwise the maximum noise I measured was -79dB. In a typical setting on a vocal, and at the 50dB setting, I set the gain at "2", about 40db less gain than max.
The noise doesn't seem terribly offensive so far, mostly 1/f noise with a corner frequency of 125HZ.

Now, the mods.

The circuit mod as proposed by Ed Evans changes the main feedback resistor (R14, 220k) by adding a 15k resistor in parallel to it, making a total resistance of 14k. This reduces the gain by 12.4dB and makes the operation in a +4 world more manageable. It does this by increasing the negative feedback. Increasing the negative feedback in an amplifier raises a red flag for me. I wanted to know what this mod would do to the transient response and the balance of the harmonic content of the distortion in the circuit.

My initial findings are what you would expect, in the original configuration the distortion is largely 3rd harmonic and the higher harmonics increase smoothly as the signal level increases. This makes for a nice fat dynamic character. The modified circuit tends to increase the higher order distortion products more abruptly. This tends to sound more raspy. you certainly know when the circuit is clipping!

As an alternative to the Ed Evans mod, you might want to pad the output by making a voltage divider of three resistors connected in series across the transformer secondary , a 270 ohm - 100 ohm - 270 ohm. The output XLR connects across the 100 ohm in the center... This will allow the output transformer to saturate a bit too, if you like that sound.

 

[JLM Audio]

The output cap C5 is used as the boot strap cap as well to turn up the the bias current in the output pair Q5 & Q6 as the output level increases. But this also puts some DC into the primary of the output transformer. So the same idle current is added out of phase into the other primary winding to cancel any core saturation which would reduce output headroom.

The main mod for LA3/A is to reduce the gain by making R14 smaller which reduces the noise floor and improves the amp THD & IMD greatly. I have seen values as low as 1.5k which makes the output amp basically unity gain.

Joe