G-PULTEC problems, please help!

[gyraf]

I'm seeing a -3dB point of maybe 95 Hz? Too high, should reach down to ca. 25Hz at least

What transformers are you using? Driven by what source impedance?

Amplifier output coupling capacitors?

How is your frequency response BEFORE the output amplifier (so that we know where it messes up)?

/Jakob E.

 

[lars on]

Hello Jakob,

I don't know exactly why but today the frequency response is better. ;D
Maybe because the connection of the (bad) header from the filter board on the main board wasn't good...

What transformers are you using?

By the way, my power transformer is from Don-Audio. And the audio transformers are the Lundhal LL5402.
As you did.

Amplifier output coupling capacitors?

The coupling cap (if you mean the big 4,7uF/250V ) is a big red WIMA MKP4

Driven by what source impedance?

The signal is coming from my RME and it's a 75 Ohm output.

How is your frequency response BEFORE the output amplifier (so that we know where it messes up)?

Check the curves BEFORE and AFTER the output trafo

 

[lars on]

I know it's less than one db difference but I wonder from where this bump and hollow are coming from.

 

[gyraf]

Looks correct now. The bump is transformer inductance vs. coupling capacitance, you can tune vith out capacitor value if you like, but 4u7 and LL5402 sounds pretty nice.

/Jakob E.

 

[groselicain]

Please forgive me if I'm asking a silly question. Thank you so much for this project, and even more for your continued help in troubleshooting builds.

I've got two channels built that pass audio and all the filters work as they should. However, I've been trying to get the zeroed (all gain pots set to zero) output to match the bypassed output. I tried the 1k trimmer on the high cut pot and got some gain, but nowhere near what I needed to make up the difference. One strange thing that I noticed is that when I was using my multimeter to measure the input XLR AC  on pins 2 and 3 against the output XLR, touching a probe to pin 3 of the output XLR boosts the output—without touching the other probe to anything. Touching a probe in the same manner to pin 2 of the output XLR adds additional noise. Additionally, if I unsolder pin 2 from the output, my noise goes down. Here are voltages from one of the channels—both channels are nearly identical in all respects. This is measured receiving +4db from a Whirlwind Qbox, which in the real world rarely puts out exactly +4dBu/1.227 V for me.
Input XLR
Pin 2 reference to Pin 1= .646 V
Pin 3 reference to Pin 1= .594 V
Pin 2 + 3 = 1.236 V

Output XLR
Pin 2 reference to Pin 1= .004 V
Pin 3 reference to Pin 1= .651 V
Pin 2 + 3 = .655 V

AC at pin 2 of LL5402 = 1.336 V

I'm pretty confused. I'm using the LL5402 on the output. Looking at those voltages, is there some really simple problem I'm overlooking? Thank you so much.

 

[gyraf]

..for transformer balanced, you need to measure between XLR pins 2 and 3

/Jakob E.

 

[groselicain]

..for transformer balanced, you need to measure between XLR pins 2 and 3

/Jakob E.

Right, which is why I included that measurement: .655V. 

Looking at my post now, I realize you might’ve been confused by the “+”—I wasn’t arriving at that number by adding the two numbers together. That was just the way I typed it. I measured between pins 2 and 3.

Thanks Jakob.

 

[groselicain]

So, no suggestions from anyone?  Is there really a way to damage the Lundahl transformer? That’s where things seem to go wrong, but I don’t see how. I mean, you can’t put it on the board incorrectly, but I can’t see another issue. Could someone help?

Thank you.

 

[RJS]

Hi Everyone,

I just completed my second unit and everything is working except the High Boost.--I get nothing,regardless of frequency selection.

I did the basics, reflowed the solder connections, made sure the connections were correct and checked for broken traces and continutity.  Everything looks okay.  I compared the wiring and connections to the first unit and nothing stands out.

I have to be missing something.  Is there anything in particular I should check or look out for?


Thanks,
Robert

 

[gyraf]

Refer to the simplified schematics in my first post in this thread for an explanation of how high-boost works. It's basically about partial (adjustable) bypass of the "hi boost pot" with a L/C series resonator/bandpass made up by your selected cap and inductor.

..for high boost, check continuity/path:

- from input to hi-boost-pot wiper,
- to one selected frequency Cap,
- to and through a selected inductor,
- through one half of the bypass switch,
- and returning to bottom (CCW) of the high boost potentiometer..

Error-modes are here anything that breaks this signal path, like open-circuit inductors or switches or miswired pot

/Jakob E.

 

[RJS]

Thanks Jacob. 

I traced everything and it looked okay and was getting continuity.  However I tested the rotary switch and it seemed to be shorted in parts.  I looked at the other unit (the one that works 100 percent) and discovered the the continuity tests did not match between the two units.  Looks like I have a bad rotary switch.  Will replace and report back.

Thanks,
Robert

 

[missi93]

Hey guys!

Just finished building my first big DIY Project, the G - Pultec!!

It's working and sounding amazing but having a few problems that I was hoping to get some answers on.

First is the high cut. It seems to work but has a very small effect in the high frequency area, only cutting like 1-2 dB or so. I still haven't analyzed or made any graphs so I can't show you the problem but I hear it is cutting way less than it should.
The attenuation is audible for the first lowest value (33nF +33nF)  but at the highest (150nF + 33nF) I can't hear any cut.

I am using WIMA MKP4 series with values from 0,033uF - 0,15uF which I believe correspond to the nF values..

Could it be the 75R resistor? Tested the "Cut Hi" pot and it sits around 1K so nothing wrong there.

Second, my by-pass function is also acting weird. It by-passes normally the low cut, hi-boost and hi-cut but does not affect the low - boost section, which can still boost while by-pass is active.

Third, I feel like when giving maximum low boost the rest of the signal tends to "compress" or "distort". I feel like my high end gets a bit lost when boosting the low fully. Is that normal?
(at highest low boost it is though VERY LOUD low end)

I hope I made myself understandable, sorry for not having any graphs or curves. Actually, that brings me to another question. Which software are you using for analyzation? I found some online but they are only available as .exe ,  since I am a Mac user..

Thanks for your time.
M.

 

[gyraf]

re-check your wireing, there's something wrong.

Are you using a kit? Then best to ask kit maker

The attenuation is audible for the first lowest value (33nF +33nF)  but at the highest (150nF + 33nF) I can't hear any cut.

something wrong here, 33nf is supposed to be the highest freq, not the lowest

Look at the simplified schematic of filter functions I posted  halfway down on page1

I have a feeling you have a miswired bypass switch

/Jakob E.

 

[missi93]

Hey Jakob, thanks for the answer.

I will check on the bypass switch if it's working properly.
Sorry about the nF values, I mirrored them ofcourse..

I am using the pcb from the friends at pcbgrinder so there shouldn't be any connection issues.

I re-checked the hi-cut problem and figured out that it's connected to the low-boost pot.
So Hi-Cut is cutting good but only when I open for the low boost pot. Any ideas what could be causing this?

This also explains how my "high end" got lost when I boosted drastically in the low haha..

Best,
M.

 

[missi93]

Actually further testing shows that low boost works in bypass mode but only for around first 70% of the 10K log potentiometer.

So when I exceed a certain value it quickly reduces to the original signal.. (only in by-pass mode, boost works normally when Filter is active)

I checked the wiring on the By-Pass Switch and I don't think it's the switch causing it. (All connectivity was right on the switch) and there is basically one way to put the switch in so it couldn't have gone wrong.

I'm using a normal 2x6 switch with a adjustable ring to fix the amount of switches, to switch between the first two values as seen on the PCB layout.

Thanks,
M.

PS
This is a major school project for me and I will be comparing this unit to a original Pulse Techniques gear and other EQP1A clone regarding THD,  Filter Curves, Noise etc..
I ask around for any advice for analyzation software or proper Sweep and Signal Generating Devices/Softwares.

 

[gyraf]

Make sure to turn switches all way counterclockwize before inserting the locking pin?

 

[RJS]

I just wanted to follow up on my issue I was having.

It turns out that the selector switch for the high boost was defective.  I replaced it and it is working perfectly now.

Thanks!
Robert

 

[gyraf]

Thanks for coming back with solution, Robert..!

 

[missi93]

Hey I also wanted to re on my problem. Everything is working good now! It was a broken connection between first two pins in the Low Boost Pot, explaining why the circuit was completely closed for the Hi-Cut as well unless you open for the potentiometer.

I wanted to ask around though about some graphs I made from running signals through the device.

The one below is a 4dBu signal sweep through my G-Pultec showing the values around 100Hz

The device is very low noise which is perfect but it has a bit more distortion than expected, going at 0,95% THD at 100HZ

To compare I have also done a sweep through a original Pulse Techniques which has much lower distortion values

 

[missi93]

This is the same sweep but in a Pulse Techniques version. All sweeps are done at level ~1.228V. Although it seems they have different amplification since the signals are coming back with a few dBFS difference.


I wanted to ask about what would you think is causing the distortion in my G-Pultec? Although it really doesn't matter in my opinion it sounds very musical and massive