gpultec low end linearity improvement

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

atticmike

Well-known member
Joined
Oct 28, 2011
Messages
502
Hey there,

Ian was so kind enough and popped for a LL88 ASC transformer, I'm really grateful for that :)

Now, you guys know how I solder the windings to the gpultec board? The winding informations are as follows:

"Blue/Yellow are the primary winding with White as a centre tap, Red/green are the secondary winding."

Help's most appreciated and thanks in advance :)

*Update* The malfunction did not occur from the input transformer but the amplifier / psu.

*Update 2* I dedicated this thread now to optimize gpultec's linearity at low-frequencys.
 
Probably no one has any idea what this is...  First 2 results in google tell me it's a 1:1 trafo, so you're going to use this on the input.  Isolate the CT, don't connect it anywhere, and connect the primary to the XLR 2+3 and the secondary to the input on the PCB.

hmm, here's some polarity info:

http://www.groupdiy.com/index.php?topic=47901.msg603991#msg603991

On second thought, the Gpultec uses PCB mounted lundahls I think, so you can connect everything as usual and just connect this trafo to the PCB minding the Lundahl pinout.
 
mitsos said:
Probably no one has any idea what this is...  First 2 results in google tell me it's a 1:1 trafo, so you're going to use this on the input.  Isolate the CT, don't connect it anywhere, and connect the primary to the XLR 2+3 and the secondary to the input on the PCB.

hmm, here's some polarity info:

http://www.groupdiy.com/index.php?topic=47901.msg603991#msg603991

On second thought, the Gpultec uses PCB mounted lundahls I think, so you can connect everything as usual and just connect this trafo to the PCB minding the Lundahl pinout.

I just hoped that if I posted what the primary, secondary winding and center tap was,  someone would be able to tell me how to hook the transformer up. Because that's what it's all about, ain't it? And going by the documents provided by lundahl, there is no such description.
 
welll, it's not that hard really. You should look a bit more at this PCB until you understand it, it will help you going forward.

look at this and try to see how jacob initially intended the trafo to be hooked up:  http://www.lundahl.se/pdf/5402.pdf

Anyway find the input trafo on the amp PCB. There is a 10K resistor right by it, put 600R there, not 10K. 
Primary of the trafo goes to pin 4 and 5, secondary goes to 10 and 3.  If polarity is reversed, just flip the wires on either primary or secondary.

 
mitsos said:
welll, it's not that hard really. You should look a bit more at this PCB until you understand it, it will help you going forward.

look at this and try to see how jacob initially intended the trafo to be hooked up:  http://www.lundahl.se/pdf/5402.pdf

Anyway find the input trafo on the amp PCB. There is a 10K resistor right by it, put 600R there, not 10K. 
Primary of the trafo goes to pin 4 and 5, secondary goes to 10 and 3.  If polarity is reversed, just flip the wires on either primary or secondary.

How do you figure out swapping 10 for 600r? In this would be ideal for the new trannie?

Going back to running this build with the lundahl, you're thinking I got it hooked up at the wrong input?

redcircle.jpg


Was I supposed to connect it to the adjacent input unbalanced?
 
LL88/ASC is a BBC part number. I doubt whether Lundahl have anything to do with their manufacture, I think in this case 'LL' denotes 'line/line transformer'.

These were typically used by the BBC at the 600Ω output of transistor line amps for galvanic isolation and line balancing, although the CT does suggest another function, such as maybe a general purpose microphone input with dual impedance (most of the BBC push-pull circuits have tertiary feedback windings).

 
Damn it. Just took out the input transformer (that lundahl out) and nothing has changed, even without the input trannie, sill that drop.

Could it refer to ground and out being connected?

I've fucking built like 10 things already and some of them were beyond difficult but this, this goes beyond my perception of the average rationality.

First unit I ain't get working is a fucking passive pultec. I'm gonna burn this at the next Rock festival.
 
MagnetoSound said:
LL88/ASC is a BBC part number. I doubt whether Lundahl have anything to do with their manufacture, I think in this case 'LL' denotes 'line/line transformer'.

These were typically used by the BBC at the 600Ω output of transistor line amps for galvanic isolation and line balancing, although the CT does suggest another function, such as maybe a general purpose microphone input with dual impedance (most of the BBC push-pull circuits have tertiary feedback windings).
I understood he's replacing the (normally used) lundahl with the BBC trafo he got from Ian (ruffrecords?).

Anyway, I saw your other thread about the hpf, and I don't think it's your input trafo.  I dont have an aurora, but I have a few (solid state) pultecs, and they are all flat at the bottom. The (filter) circuit is not much different from the Gpultec.  Mine actually have a slight rise in the end similar to the rise in a API 312 preamp (all mine now have 2520-based amplifiers). 

Have you tested the amplifier by itself?  Just put a jumper from IN -> Out on the amp PCB. (two connector to the left of the one you circled in the image you posted) and connect the aurora to your XLRs like before. 

To test without the input trafo (direct to tube amp) just connect the aurora output to the unbalanced in (connector immediately to the left of the one you circled)

To test the input trafo by itself, connect the aurora output to the balanced in (circled connector) and the aurora input to the unbalanced input of the PCB.  (leave filter disconnected for the above tests).

Don't get discouraged, pultecs are (usually) very easy to build, but you have a mistake somewhere, or faulty component, or cold solder joint, or whatever.  I have one unit I got from a friend who couldn't finish it, and I still can't get it working. In the end I think I'm going to just get a new PCB and transfer everything over, it'll be faster than trying to find the problem (many cold solder joints up til now).

If you insist on trashing it, I'll send you my address. Let me do the trashing for you!  ;D
 
alright, that's what it looks like if I connect straight in and out without hitting the filter board:

straighinandout.jpg


unbalanced in (removed the input trannie, did put cold and gnd at the ground in and hot at +, as instructed on the pcb.

shouldn't the signal be all flat?

the frequency response looks like a f**king mess.

The only custom modification I did was to ground the cover of the tube since the ground area is right below it (shielding).

The parts I used are as follows for the power board:

- regular wima big cap  WIMA MKS4 4,7uF, 250V LS27,5

- JJ 220uF, 400V radial

- Panasonic NHG 4700uF 35V

- Panasonic NHG 470uF 35V

- Panasonic NHG 10uF 160V

- WIMA MKS4 220nF, 250V LS7,5

- B380C1500

Here's a picture of the board:

redcircle.jpg


I got no idea where that's coming from.

Oh and that's the tube:

http://www.ebay.com/itm/370334270873?ru=http%3A%2F%2Fwww.ebay.com%2Fsch%2Fi.html%3F_from%3DR40%26_sacat%3D0%26_nkw%3D370334270873%26_rdc%3D1
 
mitsos said:
Did you try the amplifier with the input trafo, going balanced in?

I could put the lundahl back in and try it that way. I'll let you guys know how that worked out.
 
alright so I put the lundahl back in, swapped to balanced and get the same response as before...

I mean, what could seriously be now the issue?

Faulty tube, capacitor, resistor, rectifier?
 
Don't know, but before I lead you on some wild goose chase, maybe post some pics of the PCB top/bottom in the main Gpultec help thread and see if anyone spots something obvious. It's usually some small thing overlooked. 


 
Are you following G-pultec schematic for all part values? if so lower the input transformer secondary load 10K resistor, 1K (or something close) in parallel with the 10K that is already there and then check for improvement.
 
eliminated the tube cover grounding by removing the shell without any improvement, gonna try the resistor swap asap

response.jpg
 
Check the 4u7 capacitor. If for some reason it is of too low value, there would be a high pass filtering effect like that.
 
noulou said:
Check the 4u7 capacitor. If for some reason it is of too low value, there would be a high pass filtering effect like that.

that's the one I got

WIMA MKS4 4,7uF, 250V LS27,5
 
Here are couple of pictures. I swapped the 4,7uf capacitor by the way with a new one and nothing's changed. Will it really do anything if I swap the 10k for 1k?

pultec1.jpg


pultec2.jpg


pultec3.jpg


Doesn't anybody have a clue what might go wrong with it? Gyraf?
 
As I see it, you have a -3dB point of ca. 1500Hz

A quick calculation - e.g. http://www.muzique.com/schem/filter.htm shows that assuming a 220nF input capacitor at the filter board, you are likely to have a input resistor of ca. 470 Ohm.

Could it be that you by accident have a 470R at the 470K input resistor position (or possibly swapped the two)?

Jakob E.
 
Back
Top