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

[SmokeyJoe]

Yeah, but check the prices of these trx, you'll be surprised, and the quality seems to be really good. Check this thread UTM. I've ordered a pair of inputs and output transfomers for my dual REDD47, and they didn't take long to arrive, but I'm from France.

Thanks for the link and recommendation, I'll give that a read.

The trouble is, since we left the Customs Union it's been unclear if I'd be charged import duties. However, after some research it looks like it may not be as bad as I thought. Although the UTM prices likely makes it worthwhile either way.

Why not ordering from AML since you're in UK? They have both Carnhill and Sowter I believe and you won't have to mess with import taxes, etc...

That's a new one to me, cheers for the recommendation.

 

[SmokeyJoe]

Yeah, but check the prices of these trx, you'll be surprised, and the quality seems to be really good. Check this thread UTM. I've ordered a pair of inputs and output transfomers for my dual REDD47, and they didn't take long to arrive, but I'm from France.

I read through the thread, looks like Igor's transformers have been really well received, and the price is definitely right. Unfortunately I can't see a suitable 600:600 output transformer. Any suggestions would be much appreciated.

Looks like I can get VTB2281 and VTB9046M from AML for under £200, and I think I'll struggle to do much better than that.

 

[dmp]

If the price is close I would definitely go with sowter. High quality iron. You won't regret them.

 

[SmokeyJoe]

If the price is close I would definitely go with sowter. High quality iron. You won't regret them.

That price is for Carnhills, which I understand are decent quality. It turns out suitable Sowters are nearly double that, unfortunately (1460, 1461).

 

[SmokeyJoe]

Thanks for the help with this so far.

How tightly should components be matched between channels for a stereo build? Are 1% resistors and 10% poly caps good enough? Should the Zener pairs need to be matched between channels?

I'm also trying to figure out if the resolution can be increased on the input/output rotary knobs to say, 0.75dB steps. I realise this would be at the expense of range if the number of steps is not increased. Anyone know if this is possible?

 

[fripholm]

Are 1% resistors and 10% poly caps good enough?

Yes. For electrolytics even 20% should be fine.

Should the Zener pairs need to be matched between channels?

Couldn't hurt.
Remember, you won't be able to get the channels matched perfectly with regards to other components. Be ready to embrace a bit of imbalance

...say, 0.75dB steps. I realise this would be at the expense of range if the number of steps is not increased.

Exactly. If you had a 10 position switch you would have 7.5dB of overall gain in this case - or 15dB if you had 20 steps etc.

 

[SmokeyJoe]

Thanks for taking the time to reply fripholm.

Yes. For electrolytics even 20% should be fine.
Couldn't hurt.
Remember, you won't be able to get the channels matched perfectly with regards to other components. Be ready to embrace a bit of imbalance

Ok, I can live with that!

Exactly. If you had a 10 position switch you would have 7.5dB of overall gain in this case - or 15dB if you had 20 steps etc.

So to be clear, the total value of the pot/stepped attenuator doesn't need to remain the same as spec'ed (10k)?

If I reduced it to 5k total by reducing the resistor values, I could decrease the total gain and step size? Would that be the case for both input and output?

 

[fripholm]

If I reduced it to 5k total by reducing the resistor values, I could decrease the total gain and step size? Would that be the case for both input and output?

Unfortunately it's not that simple. The attenuators are in parallel with their respective inputs which in return would reduce the input impedance with e.g. your hypothetical 5k input controls. You don't want that, especially between compressor and output stage.

And I don't think the resistor values would scale accordingly. I've come up with the values from the guide using LTSpice and and a whole lot of trial and error. There may be some (complex?) math behind this but this is way over my head

If you want finer control, use pots and/or control your levels at the previous/following stages. FWIW, I've never felt like needing finer control with my units but YMMV...

 

[SmokeyJoe]

Unfortunately it's not that simple.
...
If you want finer control, use pots and/or control your levels at the previous/following stages. FWIW, I've never felt like needing finer control with my units but YMMV...

Ah ok, I misinterpreted your last message. Thanks for the explanation.

What is the pitch of the board connectors? Are they JST XH (2.5mm) or something else? I can see some larger connectors too, what did you use there?

I'd also like to confirm the wiring gauges if possible please, as I can't tell what current capacity is required for different parts of the circuit.

 

[fripholm]

The pitch of all the connectors is 2.54 mm. There are no larger ones - unless I'm not really getting what you mean.

Use wire which you can comfortably work with. There are no special requirements with regards to current capacity.

 

[SmokeyJoe]

The pitch of all the connectors is 2.54 mm. There are no larger ones - unless I'm not really getting what you mean.

Use wire which you can comfortably work with. There are no special requirements with regards to current capacity.

Ok thanks, good to know.

Over in the sale thread you posted an internals photo and I could see you've used chunkier connectors and thicker gauge wires in places, so I wondered if that was a requirement. Sounds like it doesn't matter though.

 

[fripholm]

No, it doesn't matter. This is what I had available at the time so it ended up in the builds

 

[wrentema]

Thanks for all of the answers!

Okay, that makes sense. The way that I am implementing Auto-release and Hold is on two separate, interdependent toggle switches (a DPDT and a 4PDT). Auto-release will be on the DPDT switch, just like it is meant to be. When I toggle the 4PDT Hold switch "on," the connections for Auto-release will be broken (R51 will not connect directly to ground OR to C32/R90, but rather to the Hold pot wiper and the anode of the added 1N4153 diode) and the connections for Hold will be made, and the auto-release switch will no longer function. In this way, I am making it so that I can use either Hold or Auto-release, mutually exclusive to one another, at any given time, OR opt to have neither one engaged. Don't know if this description was clear, but I'll share my wiring diagram at some point when I have a chance.

The main reason I'm doing it this way it two-fold, and somewhat silly perhaps -- first, since the build guide presented Hold and AR as either/or options, I got it in my head to think of them as such, and that notion just kind of stuck; second, totally for vanity's sake, the proportions of my front panel design look a little better with a couple toggle switches breaking up some of the empty space! ;D Your answer was very helpful in confirming how I've planned this wiring, thank you!

Did this work out? Do you still have this wiring diagram somewhere?
wondering what to do with the Hold/Auto release thing.. Definitely want hold.. Both would be ideal of course..

And did people match transistors between left and right? If yes.. what do you match.. I guess hfe right?

Starting my build soon. Read the whole thread here (took a while!). Thanks for all the great info here.

Now fingers crossed

 

[Bernardo]

Hello colleagues, how are you?
After months of intermittent construction for work and family, I have finished the Tg zener limiter compressor project.
Everything seemed to be going well, in the construction I was testing the device with an unbalanced signal from the mobile.
But once I tried it with digital converters with balanced output, things changed.
I have the following problems.
-at three quarters of the input potentiometer the signal begins to saturate, the compression is triggered.
-When using the hpf filter it starts to saturate in the same way as when I force the input level.
-thd mode is impossible to use it distorts a lot.
-the output level is extremely lower than my other dynamics processors.

I have done all these tests with and without audio output transformers, and they saturate and distort the same.
The zener diodes are matched purchased from a fellow forum member.
The unit is quiet, but I have this distortion problem when I go past the middle of the input pot.
Please any clarification or help will be appreciated.
Thank you very much.

 

[fripholm]

More info please.

Where does the distortion happen, compressor or output stage?
What were the signal levels when using the "mobile" vs "digital converters"?
What's the ratio of your input and output transformers?
Do you have the same problem on both channels?
Did you make all the calibrations with a possibly weaker signal from the "mobile"? Did you calibrate at all?

 

[Bernardo]

I can configure the outputs of the converters to +4 or -10.
at -10 it hardly sounds, at +4 it is more usable, but it distorts the input as I mentioned before.
Sounds like an impedance problem.
-I can't tell if the problem is with the output stage or the compressor, I don't know how to find out.
-If I have followed the guide at the time of calibration, the FT2A trim, which according to the guide is for the output gain, I hardly notice the change when turning the trim.
-Both channels have the same problem.
-I have stopped the output tx.
- the input ones are utm2546, similar to carnhill. configured 2:1
10k:2.4k.
attached image.

 

[fripholm]

Rule out the output stage by injecting a signal at the center pin of BYP2 (disconnect anything else from this connector) and checking the output. Is it clean?

Example

What do you mean by:

-I have stopped the output tx.

Are you sure you followed the instructions about installing FT2A correctly? Here's the snippet about the trimmer copied and pasted from the guide:

In my experience, adjusting R22 is not needed most of the time. Just install R22 with
its default value of 12k and solder a jumper from trimmer FT2A’s squared pad to its
middle pad.
If you want to give it a try, use dedicated trimmer FT2A – 20k. Install a 10k resistor for
R22 and adjust the trimmer as described below.

It basically says, don't install it if you don't have a particular reason to do so.

 

[Bernardo]

What do mea by:

I have stopped the output tx

I have removed the output transformers to make sure it is not the problem, when the unit is well adjusted I will install them again.


First of all, thank you very much for your time and, you have always solved all kinds of doubts.
I have some doubts about what you say.
-for Ft2A you have to install R20 (12k) and make a pin in the drawing of the trimpot?.
-For Ft2b, resistance 56 is installed and the trimpot drawing is bypassed?

 

[Bernardo]

Rule out the output stage by injecting a signal at the center pin of BYP2 (disconnect anything else from this connector) and checking the output. Is it clean?

a lot of distortion

 

[fripholm]

-for Ft2A you have to install R20 (12k) and make a pin in the drawing of the trimpot?.

Yes, omit the trimmer, install R22 (and R20 as well, which was a typo on your side I suppose ) and connect the squared pad with the center pad of the trimmer. Exactly as explained in the guide.

-For Ft2b, resistance 56 is installed and the trimpot drawing is bypassed?

FT2B on the other hand is to be installed for a more convenient calibration. Omit R56 in this case.

Just to clarify: both trimmers FT2A and FT2B were added as a convenience but it turned out that FT2A wasn't really needed (hence it's marked as optional in the BOM) as opposed to FT2B which makes calibration much less complicated. I thought that was clear from the guide.

a lot of distortion

Huh?? So it's the output stage?!

Is it the same distortion you're hearing when you were doing your initial tests described in this post?