[BUILD] fripholm's TG1 Zener Limiter boards - support thread

[weiss]

Hi andow,

no pm received here. I have to check if there are some left.

i would be also interested!!

 

[weiss]

hey, ive finished the beast and to my ears it sounds great 

some notes:
lundahls in/out -  imho recommandable
putting the PSU external eliminated all PSU related noise (eg 60hz hum)
zener diodes from RND components are really low tolerance zeners (ive got matched ones pretty fast)
soundwise its an agressive compressor with alot of color/thd, similar to rev A 1176 but more classy sounding

thx fripholm!!!

here's some audio:
https://www.dropbox.com/s/lom2ky1wh3imow2/zener%20drums.mp3?dl=0

perhaps ill order a cnc engraved panel, mine looks very DIY

nice build. audio sounds great! hope to finish mine soon 

 

[fripholm]

It's a nice build indeed. Regarding the front panel, I wouldn't change a thing - looks great

putting the PSU external eliminated all PSU related noise (eg 60hz hum)

Here's a quick tip to reduce hum induced by internal power transformers:

These are the lams of a 300W power transformer from an old halogen string light that I 'reverse-engineered' a few years ago. The lams are similar to the transformer shields you can buy at e.g. Don Audio. Those work fine but are insanely overpriced IMHO.

I just wrap two or three layers of these lams around my internal power transformer and secure it with electrical tape. Both of my TG units have it as well. Here it is in a 6-channel mic pre that I designed:

Place the transformer as far as possible from audio circuitry, wrap the lams around it and rotate till the hum is gone. HTH

 

[andow]

I'm in the middle of wiring my Zener Limiter and I'm wondering why there is no pad on the PCB to connect the input ground?

 

[fripholm]

I'm in the middle of wiring my Zener Limiter and I'm wondering why there is no pad on the PCB to connect the input ground?

The input connector's pin 1 (if using XLR) usually goes directly to chassis ground which is then connected to PSU GND and from there to the power supply connector on the limiter board.

As I have the I/O transformers on a separate (perf) board, these transformers share the same ground connection coming from the output connector on the main PCB.

 

[andow]

The input connector's pin 1 (if using XLR) usually goes directly to chassis ground which is then connected to PSU GND and from there to the power supply connector on the limiter board.

As I have the I/O transformers on a separate (perf) board, these transformers share the same ground connection coming from the output connector on the main PCB.

I usually take the PCB ground directly from input pin1 (which is of course directly connected to chassis).

Ok, it makes sense if you have in- and output transformers on the same board. I don't, so an extra ground pad at the input would have been nice. No big deal, maybe something to consider for the next revision... 

 

[andow]

I finished mine today and everything seems to work! BUT there is pretty loud noise, that also others wrote about in this thread, when using the Zeners from Mouser.
Anybody here that has built a quiet unit and has some Zener diodes left?

 

[Potato Cakes]

Which Zeners are you using? I went through several batches from different manufacturers/vendors and all of the 1N751s that I tried were way too noisy. I swapped them out with 1N5231s and the noise is a non issue now, and I didn't have order from a bunch of different sources to find some that worked. I would recommend doing the same if you are struggling to find 1N751s that have a usable noise level.

Thanks!

Paul

 

[andow]

Which Zeners are you using? I went through several batches from different manufacturers/vendors and all of the 1N751s that I tried were way too noisy. I swapped them out with 1N5231s and the noise is a non issue now, and I didn't have order from a bunch of different sources to find some that worked. I would recommend doing the same if you are struggling to find 1N751s that have a usable noise level.

Thanks!

Paul

Thanks for the reply! I'm using 1N751 Central Semiconductor from Mouser and these are unusable. Did you notice any difference in sound between 1N751 and 1N5231?

 

[Potato Cakes]

Thanks for the reply! I'm using 1N751 Central Semiconductor from Mouser and these are unusable. Did you notice any difference in sound between 1N751 and 1N5231?

With the noise being so bad with the 1N751 I didn't make any comparisons. I did note in my earlier post that this unit sounds like a similar Chandler unit I rented years ago. I don't know how it compares to an original (as I don't have one). It is a different character than the other compressors have.

Thanks!

Paul

 

[fripholm]

Did you notice any difference in sound between 1N751 and 1N5231?

Originally, I had  1N751 zeners in  the three channels I've built (I was lucky and catched a low noise variant) and eventually swapped all of them for 1N5231 diodes. They were even less noisy (~2-3dB). I didn't notice any difference sound-wise. HTH

 

[ron_swanson]

With the noise being so bad with the 1N751 I didn't make any comparisons. I did note in my earlier post that this unit sounds like a similar Chandler unit I rented years ago. I don't know how it compares to an original (as I don't have one). It is a different character than the other compressors have.

Thanks!

Paul

I purchased 10 lots of 1N751A Fairchild Zener Voltage Regulator 500mW 5.1V 70mA Glass DO-35  on Ebay for my Zener build from this store... 

https://www.ebay.com/itm/Lot-of-15-1N751A-Fairchild-Zener-Voltage-Regulator-500mW-5-1V-70mA-Glass-DO-35/192015781782?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649

No problem finding many, many matched sets within the first 30 or so tests I made out of the 150 Zeners delivered .  My build is very quiet. 

 

[andow]

Thanks for your help! I ordered another batch of 1N751. If these are not silent enough, I'll give 1N5231 a try.

 

[andow]

I finally tried Fairchild 1N751 and they are silent! 

So my Limiter is basically working and already sounds wonderful!

HOWEVER, I'm getting distortion on the first channel at high levels, the second channel is clean (in bypass that is, so the problem is in the output stage).  I'm pretty sure the output stage is unstable: I measured oscillation with my oscilloscope, and when touching the pins of the output transformer the distortion goes away.
So as suggested in the guide I tried 22R between the output of the board and the transformer but it didn't really make a difference.

Any suggestions what to try next?

 

[fripholm]

Have you tried bigger resistors, 47 or 100 ohms? Sometimes also a small ceramic cap from output to ground can help - like 100pF or so.

You may have done this already, but diligently compare the output stages of both boards for any unintended differences and - if possible - try swapping their position in the chassis to see if the other one starts to oscillate.

 

[andow]

Have you tried bigger resistors, 47 or 100 ohms? Sometimes also a small ceramic cap from output to ground can help - like 100pF or so.

You may have done this already, but diligently compare the output stages of both boards for any unintended differences and - if possible - try swapping their position in the chassis to see if the other one starts to oscillate.

Thanks for your help!
I hope to find time this weekend. I'll let you know how it goes...

 

[andow]

I found it!  ;D
I used shielded cable between the output stage and the output transformer and its capacitance seems to have caused the oscillation. I don't really know why, since it was the first channel with the shorter cable, but whatever...

After some calibration this thing sounds very nice and it's very versatile too! I can see myself using this on a lot of different stuff.
I especially like the distortion in THD mode, sounds wonderful on drums!  8)

One thing about gain staging: even when using an 4:1 input transformer I usually have the input gain in the lower, and the output gain in the upper half of the range. And when applying just a few dB of gain reduction the output stage doesn't have enough gain to achieve unity gain.
The input is not really a problem, but the lack of output gain is a bummer. Has anybody tried using a 1:2 output transformer? I assume this wouldn't be a problem since the output impedance is pretty low anyway, right?

 

[fripholm]

As you are using a 4:1 input (why?) it would make sense to use a 1:2 output transformer. The output stage is certainly beefy enough to drive the higher load.

When calibrated correctly, the compressor stage alone should have a gain of about -3 to +6dB and the output stage provides +25dB with the output control maxed.

 

[andow]

As you are using a 4:1 input (why?) it would make sense to use a 1:2 output transformer. The output stage is certainly beefy enough to drive the higher load.

When calibrated correctly, the compressor stage alone should have a gain of about -3 to +6dB and the output stage provides +25dB with the output control maxed.

Thanks again for your help! (and this amazing project, btw  )

I included a switch to select between 2:1 and 4:1 input ratio and I'm pretty sure everything is calibrated correctly. I used Elma switches for input and output gain control with the same values you provided in the guide.

I just did a test (with 2:1 input transformer): when the input control is fully counterclockwise, the compression kicks in (needle moving) at +7dBu input level. If I then set the output control to fully clockwise, the output signal is at +4dBu.
Can this be correct?

 

[fripholm]

I've just checked this against one of my channels.

With the input control fully CCW and a 2:1 I/P transformer, compression kicks in at +7dBu - as you said. With the output control maxed I'm only losing about 1 dB which is even less considering that my interface by itself already loses ~0.7dB when connecting output and input directly.

Is it possible that your interface has a different sensitivity set for the input than for the output (assuming that you're using an audio interface)?

The steps from FINE TUNING II in the guide also adjust the output level - not primarily though. If nothing else helps, you could try to find a compromise between distortion and level with those as well.

Viel Glück!