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

[thomb]

Hallo,

i am in the tuning process of this nice sound box. There are a few problems i can´t solve:

I have built in I/O Transformers, no Hold mode, Switched Attenuators for Input/Output levels. All switches work like they should so far. And it sounds nice!

1. But in bypass mode, the right channel is about 5 dB lower then the left one. I checked components and cabling in the output section.
2.  I could not level match both channels for all three modes. Once i matched  it in THD mode, in Comp mode there is a 1-2 dB difference (right channel lower).
3. I could not match the second harmonic in both channels. Whenn i match the overall levels to the same level in both channels, the second harmonic in the left channel is as high als the third harmonic, where in the right channel the second harmonic is much lower.

Maybe someone had the same issues.

Thanks
Thomas

 

[fripholm]

1. But in bypass mode, the right channel is about 5 dB lower then the left one. I checked components and cabling in the output section.

This should be the first issue to address. The output section is the only active one that the signal passes in BYPASS mode - so start by tracing the signal from the input to the output transformer through the output stage and see where it starts to differ.

2.  I could not level match both channels for all three modes. Once i matched  it in THD mode, in Comp mode there is a 1-2 dB difference (right channel lower).

Get the first step right (see above) and this might fix itself after another attempt at calibration.

3. I could not match the second harmonic in both channels. Whenn i match the overall levels to the same level in both channels, the second harmonic in the left channel is as high als the third harmonic, where in the right channel the second harmonic is much lower.

I don't know how much they are different but don't expect the harmonics to be exactly the same in both channels. IMHO, one of the reasons this unit sounds as great as it does for stereo material is because it's almost impossible to match L and R precisely. Other than that, see step 2 

 

[thomb]

Thanks Fripholm,

i traced down the signal path. When i touch the signal path after R 81, the output in Bypass mode is the same on both channels. When i leave the probe, the output gets down 5 db lower. (Measured with one end of the DMM in ACV Mode to signal path, the other probe to chassis gnd).
Which Element could cause this problem?
Maybe one Element burnt, because at the time of the first run, all BD140 and BD139 soldered backwards.  But after correction , i am getting  signal out of all channels.
Maybe Q 14 is the prospect (or components around Q14). At the left channel the voltage of the  base of Q14 is 0,47 DCV, at the right channel only 0,13VDC (Signal 1kHz sine wave -5dB at the both inputs for all above measurements).
Does anybody have an idea?

Thanks Thomas

 

[fripholm]

The bases of Q14 are at about -490mV (negative!) on both channels of my stereo unit, its collector (which is also the base of T15) sits at -9.6 volts with respect to GND.

Your left channel seems about right (if that was actually a negative reading) - the right channel is not. Try changing Q14 and/or T15 and check component values around these parts, solder joints etc.

 

[thomb]

Thanks for the fast response!
Changing of Q14 and T15, resoldering the Components around this area didn´t solve the loss of 5 dB in Bypass.
One Difference i noticed by tracing down the signal path is a Difference after R 87 (left Channel 107 mV AC, on the right Channel 139mV AC).
I am running out of ideas.
Maybe i should change the Output Transformers?
It´s strange: By touching the Output Square pad with the DMM in ACV Modus (other Probe to GND) the Voltage jumps to the same level as the good left Channel (1,9.. VAC).
Thanks for your help!

Regards
Thomas

 

[fripholm]

Changing of Q14 and T15, resoldering the Components around this area didn´t solve the loss of 5 dB in Bypass.
One Difference i noticed by tracing down the signal path is a Difference after R 87 (left Channel 107 mV AC, on the right Channel 139mV AC).
I am running out of ideas.
Maybe i should change the Output Transformers?
It´s strange: By touching the Output Square pad with the DMM in ACV Modus (other Probe to GND) the Voltage jumps to the same level as the good left Channel (1,9.. VAC).

Have you tried a series resistor (< 100 ohms) between the output and the o/p transformer or a small cap (try 22p for starters) from output to GND? What are your DC (!!) voltages after the transistor swap at the base and the collector of Q14?

Try injecting a (low level) signal at the middle pad of BYP2 without anything else wired to this connector. Is the output level the same on both channels when you do this?

 

[thomb]

Injecting a signal (just pin 2 of the XLR Signal generator) direct to the middle pin of BYP 2 results in the exact same signal at left and right output. The difference in Level does also exist, when i insert the signal in middle Pin of BYP 2!! Sorry, second measurement with little higher Levels made this clear.
22p from Output to GND doesen´t equal the levels (the same with 10P and 100P). Also with the 82 Ohm resistor in series from Output Pad to the Output Transformer.

Q14 DC Voltage on the right (lower output) channel: B = -103 mV DC, E = -10,3 V DC, C = -0,003V DC.
Q14 DC Voltage on the left (normal output) channel:    B = - 490 mV DC, E = - 10,45 V DC, C = -0,001 V DC.

But when i touch the Output Pad  via a crocodile clip (just crocodile and my Hand), there are the exact same levels on both outputs!
And the Base of Q14 has jumps to -470 V DC.
What could cause this behaviour? Defaultive BD´s or C 22, C 24, C29?
Before i desolder those Components, maybe there´s an idea?

Thanks a lot
Thomas

 

[electrisizer]

another idea: these "platinen steckverbinder" have sometimes a contact problem. i had to solder the wires direct to the pins. it was a problem on 2 and 3 pin "steckverbinder". check them for a tight contact

 

[thomb]

another idea: these "platinen steckverbinder" have sometimes a contact problem. i had to solder the wires direct to the pins. it was a problem on 2 and 3 pin "steckverbinder". check them for a tight contact

checked pin connectors a lot of times. They are o.k.
Thanks anyway

 

[fripholm]

You've mixed up the collector and emitter of PNP transistor Q14. It's really the emitter that's going to GND with your reading of zero Volts.

Anyway. Might be a bad part, wrong component value/orientation, bad/cold solder joint, leaky cap - really everything...

At least, you've narrowed it down to the output stage. Now (I'm pretty sure you've done this already) double - no, quadruple check everything with a magnifying glass, retouch joints etc.

 

[JMan]

Hey guys, I’m really putting the cart before the horse here, because I’m going to have so many dumb basic questions pretty soon, but in hopes of currying some favor ahead of that, I’d like to (hopefully) contribute something useful. 

I have been going back and forth about whether I will ultimately want to incorporate the Hold function.  Opinions on it seem to swing both ways, and I don’t want to be committed to something that I end up disliking.  On the other hand, I don’t want to miss out.

So I spent some time thinking and I believe I’ve come up with a way to have the option of both auto-release AND hold usable, switchable between the two using a 4pdt switch.  I’m kind of nervously anticipating that someone will tell me I missed something obvious and this won’t work and I’m an idiot (I already know that last part), but I figured I’d put it out there and see what you think.

I drew it out because it helps me to visualize things.  The wire from the CW side of the Hold pot could also be connected directly to the ground pad of R58, but I thought this option seemed a little tidier.  What do I know, though.

 

[fripholm]

At a quick glance this looks about right.

About a year ago I discussed this with another user and he came up with a way where you could have both Auto-Release and HOLD at the same time. I didn't try it, though.

The GND ends of both C32 and R90 need to be connected to R51, together with the newly added diode and the HOLD pot's wiper. This way the still-switchable Auto-Release should follow the HOLD voltage, too.

 

[BerndVP]

Hi,
I'm looking at the BOM, the '1N751A' I can't find on mouser, where can I get them ?

 

[fripholm]

...Reichelt in Germany has them as well as Farnell, Digikey, RS-Components etc.

Be aware, that a wrong choice of Zeners can be THE main source for problems with this build - noise, oscillation, erratic behaviour in general.

To be sure, order a couple of batches from different manufacturers - most of them are dirt-cheap.

My three limiter channels started with 1N751A (Multicomp) from Farnell but eventually I replaced them with 1N5231 from Mouser. Nothing wrong with the first choice but the new ones were slightly less noisy and they sounded identical. Others weren't so lucky - you'll see this when looking through this thread.

 

[BerndVP]

Hi,
thanks for the info, which ones did you bought from Mouser then ?

 

[fripholm]

which ones did you bought from Mouser then ?

100 pieces of each of these:

https://www.mouser.de/ProductDetail/610-1N5231BTR
https://www.mouser.de/ProductDetail/512-1N5231C

I don't remember which brand made it into the actual units as I matched a few pairs, soldered them in and they worked right out the gate so I didn't even bother to try the others.

 

[JMan]

Is there any reason (beyond front panel space) that the Auto Release connections on the push-pull switch couldn't be outsourced to a separate spdt toggle switch?

 

[fripholm]

Is there any reason (beyond front panel space) that the Auto Release connections on the push-pull switch couldn't be outsourced to a separate spdt toggle switch?

No.

 

[JMan]

1295 for the input and 1365 for the out.

I’ve returned to thinking about the transformers, and I like the idea of these two from Sowter.  (Like you, I am trying to make this a special build, but I’m working with what limited knowledge I have).

So with the odd ratios on these, were there any special considerations that you had to account for — for example, did you have to adjust R86 and R87 for unity gain in bypass?  And if you used a stepped switch for input output, did you have to change any values from what Fripholm suggests in the build guide?

Thanks for any help, I’m really working hard to wrap my non-technical brain around this build and I appreciate everyone here so much.

 

[BerndVP]

JMAN, keep in mind, Sowter is a long wait time for the moment.
I ordered 4 specific Sowters for my D-LA2A Build in May and today got the news it's delayed again for 2/3 weeks.