All things LA2a related

[alexc]

Series - near as possible to 65 or so volts with 2x 32V or 2x33V zeners

 

[canidoit]

Series - near as possible to 65 or so volts with 2x 32V or 2x33V zeners

How about series in 1 X 32 volts and 1 X 33 volts? Is 65 volts unacceptable?

 

[alexc]

Should be fine.

 

[simonsez]

I thought i'm finished with this animal, but this afternoon i put it on analyzer,  i saw a lot of harmonic dist. when i turn up peak reduction, with or without optical device (t4b). It's looks like it happen when grid V1 start to connect with grid V3 (peak red turn up). Any thought?

Thanks..
Simon.

 

[millzners]

what are you using to do the analyzing?  I'm curious for my upcoming build.

 

[simonsez]

Voxengo SPAN v.1.9.1

 

[simonsez]

No one? ??? ???

 

[SSLtech]

I thought i'm finished with this animal, but this afternoon i put it on analyzer,  i saw a lot of harmonic dist. when i turn up peak reduction, with or without optical device (t4b). It's looks like it happen when grid V1 start to connect with grid V3 (peak red turn up). Any thought?

Thanks..
Simon.

the sidechain amp section is distorting, and you have some capacitive coupling between the drive path to the EL panel and the signal which goes to the opto.

A ceramic disc capacitor can do this, as can poor layout/routing of the wires.

If anyone wants to see this for themself, they can make a 1-foot bare-wire octal extension lead, and just move the wires close together... you'll HEAR the nastiness, never mind measure it.

Layout is CRITICAL in this section.

 

[SSLtech]

EDIT: Replacing the neon lamp with a 60V zener diode (as advised by Keith in the post below) completely fixed the problem. It also tightened up the meter tracking. Voltage measured 61.8VDC across the zener when the meter was zeroed with R4.

Cool, -I'm glad that my description (a logical guess) of the 'needle falls immediately' symptom was an identifier.

I just had a neon go noisy in one of my two. Not only noisy, but flickery also. When it flickerd bright-dark-bright, the needle would kick up or down by about a dB. -This gave the user serious doubt about the signal level, even though it only affected the DC used for the GR display 'simulation' and not the actual signal itself.

I put a diode in place of the zener, and things are now gorgeous. -The benefit of the more rock-solid voltage lock was a bonus!

-As it was, your VU meter was useful as a 'relative change' indicator for AC voltage at the wall-socket.



Putting a skeleton variable resistor in place of R2 will allow you to tighten the meter tracking even further. -The reason it was so far out was that the voltage had risen so high (wiht the neon open-circuit) that the meter zero control had to be moved a LONG way out of its nominally 'correct' position... thus the values of the two sides of the 'Wheatstone bridge' were a long way off from the optimal range, for which R25 is typically set.

There's enough things I've picked up over the last few years to 'improve' the performance of 95% of the LA-2a's out there!

-God, I feel like I want to build another one, now!!!

Keith

 

[SSLtech]

Oh, and regarding Zener voltages, anything in the 60V to 70Volt range is just fine. Simply tweak the value of R25 (as you tenchically should do anyway) for best meter tracking in the ~4dB-6dB of GR range, and you're golden. -If you want to match the voltage numbers on the chart, go for the exact number, but it's only really for guidance, and about 10% tolerance either way will still work just perfectly.

Keith

 

[simonsez]

Thanks Keith!! I don't know that distorted sidechain amp has impact to the audio signal path. I will take a look closely, thanks..

 

[SSLtech]

What happens is that because it tends to clip without the limiting action of the opto, and it presents a nonlinear load when it IS present, you'll see sharp edges in the current flow or the waveform voltage.

it shouldn't leak into the signal path, but it can and sometimes does, specially if the wires run close to each other. (I just had this with one recently)

Keith

 

[drpat]

...

 

[simonsez]

After re-arrange a few wires, i can reduce the distortion. I read that LA2A can have  up to 0.5% THD!! That's huge isn't it?. Below is the harmonic dist. in about 15db GR of my LA2A. Is this normal?

 

[SSLtech]

After re-arrange a few wires, i can reduce the distortion.

Good. -That looks like ite's definitely the source of your problem. If you want to TOTALLY eliminate it, try shielding, but BEWARE of capacitance... LOOK AT THE CAPACITANCE SPECS of any coax you use

I read that LA2A can have  up to 0.5% THD!! That's huge isn't it?.

by modern standards pehaps... Back then 1% was considered quite okay  ...but it IS part of the unit's charm, so don't go nuts

Below is the harmonic dist. in about 15db GR of my LA2A. Is this normal?

Dominant third I think is commonplace, and if the unit doesn't distort with NO T4 module fitted, and NO gain on the peak reduction control, your signal path is as clean as it will get.

Whos T4 are you using? The LDRs contribute much to the distortion even with no GR applied, hence measure the distortion of the unit with T4 removed... swap LDRs to see if one is better than the other.

Keith

 

[simonsez]

I just find out that this unit is very sensitive, everything before V1 is very very sensitive. I can totally eliminate oscillation only by shielding a few wire (without grid resistor). But now you tell me that i should aware about coax capacitance too... :'(, overall it's sounds great, best for vocal track.

Distortion still leak in the signal path when i crank peak reduction (without T4) even when i disconnecting C11, I make my own T4 module with "unknown LDR" and EL panel, I will try shielding a few more wire, hope can improve more.

Thank you Keith, you're very helpfull indeed...

 

[SSLtech]

Distortion still leak in the signal path when i crank peak reduction (without T4)

That's EXACTLY how a real LA-2a behaves also, so don't worry about it.

Most people never have the T4 unplugged and the reduction turned up, since it can't do anything useful like that, but don't sweat it.

Under gain reduction yours will ALWAYS distort, as will everybody else's.

You set it for clean path with no T4, be careful with the routing and you're pretty much done.

I once built a home-made T4 which was the problem... try REVERSING the wires to the EL panel. and KEEP WIRES SHORT inside the panel... crosstalk between them cause what you describe also.

Keith

 

[simonsez]

I'm glad to hear that, coz i can't get rid distortion problem above. So i think i'm home now...
Thank's Keith...

 

[millzners]

Here's a simple question, but I've searched all through the forum and haven't found a straight forward answer:

Power Transformers:

A lot of people seem to use the Hammond 369JX http://www.mouser.com/ProductDetail/Hammond/369JX/?qs=sGAEpiMZZMv4oUrzpPKU3CkdIda8KQx%252b6DLhXp0yecs%3D which was recommended by the RecProAudio site (~$75)

Drip recommends the Allied 6K88VG http://www.alliedelec.com/search/productdetail.aspx?SKU=2270113, which I've seen mentioned a bunch here as well (~$25)

That's a big difference in price, will the Allied one really do the job ok?  I'm willing to error on the side of good performance and fewer build problems, but I thought I'd ask.  This is obviously a USA build, and I'm doing a P2P, not a PCB FWIW.

Thank you!

 

[regularjohn]

I've p2p'd 4 of these suckers with the allied 6k88vg and they all sound phenomenal.  there's no reason to pay more for the hammond.