EQP1 Dual Pultec *finished* with *Purusha Case and PCBs*

[Piedwagtail]

My suggestion is to connect every ground you have,each filter,each psu,each xlr separately direct to ONE xlr input ground pin and ground that to the chassis as close as possible ie.using the XLR socket bolt.
Star earthed-messy-but good.

Robert

 

[keefaz]

If you ground everything on ground star point, you avoid ground loop but you get bigger current loops, no ?
Eg: if you connect filter grounds on star ground point near the input XLRs, the signal have to go from line amp pcb to filter pcb and then its ground goes from filter pcb to star ground, then to line amp ground, so it takes a longer path.
For my part, I try to layout the signal ground as close as possible to its corresponding signal

 

[radiance]

Here someone with a similar problem...

http://www.groupdiy.com/index.php?topic=24940.0

 

[Piedwagtail]

Just my suggestion Keefaz.

What works is what works.Since no one had detailed star earthing I felt the builder might need that perspective.
We all know there is no answer only a number of empirical solutions to try out.

Robert

 

[hotbaby]

oh radiance!!! i think this will help me a lot as he thought about the bridge rectivers. i thought about it too but i think it is the 250v ac paralleling which causes the hum. he solved it with paralleling the dc of the new bigger rectivers... i m sure it is an ac problem as i can hear the 250 ac cables very loud if i move them about my input/output trafos! I think thats it in combination with a non effective grounding...
should i buy bigger rectivers too? 6a 600v? now i have 350v 1.5a... i think i will need bigger ones if i parallel their dc voltages to the second board... puhh there is hope

 

[radiance]

.... i can hear the 250 ac cables very loud if i move them about my input/output trafos! .

Well, I still think you'll need a bigger case. Those bigger rectifiers brought the noise levels down from
-60db to -75db, which I still find quite high...But maybe it's ok for a tube amp...dunno.

EDIT: Here's a great lecture from Bill Whitlock about grounding...

http://www.groupdiy.com/index.php?topic=23633.0

 

[dagoose]

Yep, i have/had the same problem.  :-\
I managed to get it down to an acceptable level but still it's there... I just drive it quite hard to have a better SNR and it also gets a bit better sounding when driving the tubes quite loud.
I used one big rectifier instead of the 2 parallel rects on board which helped quite a lot and then i started to work on the grouding and things like that like in the topic Radiance pointed out.
I also think the case is not big enough though... not enough space/room to really seperate the PSU from the rest of the circuit.

 

[hotbaby]

Hey Dagoose,

three days with nightmare and thinking with you all about the hum...

1. yes, removing all ac from the case will be the first step.
Usually it wouldnt be bad replacing the toroid and rectivers to an external power case. Why not. My D a n g e r o u s music machines have all external power sections...

1b. Perhaps the Elcos on both PCBs must be extern too that just linear nonpulsating current acrosses the circuit.

2. I found out that theres something wrong with the direct current (dc) if i use a blind resistor on the bottom. the problem is that the elco trys to fill the pulsating ac holes and with the blind resistors the elco is anloading to fast that theres a litle hole which increases a litle trafo hum... i will setup up the blind resistors to the power switch that it will recharge if i switch the power switch off.

3. The rest hum i think is the electric noise field of the toroid. Theres need for mumetall around the toroid and all the ac pulsating stuff.

4. In my case i rective 24vac to dc for the LED power switch near the "filter high section". The rectiver is to near to the filter pots which increases hum too. Theres need to put it more near the toroid too and shield it with mumetall. Perhaps steel will do this job too.

5. One Star grounding

6. If step 2. doesnt work perhaps we need bigger elcos which can linear the dc current better.

7. twisting the ac and dc cables.

That are the 7 Hum aspects i thought about and i am think this will help to solve the hum problem.
Remember that the Powerpcb usually is for mono use and the stereo use causes that the ac rectiving is near to the input/output trannys.
The second better think on the mono use is that there can be more space between toroid and i/o trannys and the dc will be more linear cause, i think.

(- Perhaps the gyraf pcb and suggested toroid does not load the elcos perfectly or the elcos have problems to linear the direct current if you use two power pcbs parallel in stereo mode... i must try that out with removing the blind resistors or bigger elcos)

Right now i have -65db on the right channel! I want to get it down to about -80db with this todo list!
There IS hope!

Aaron

 

[1954U1]

Interesting thread.
I'm sourcing too the parts for a D-EQP1.

If I were you, I'd follow the Radiance "external PSU" suggestion, and try what Kingston recently made for solving the hum/noise problem on his G9:
put outside of the audio PCBs, on veroboards, the PSU's components: diodes rectifiers, caps [maybe attached to the chassis], regulator.
You can at least try it on only 1 channel, disconnect at all the other channel and see if the noise improve considerably.
A spectrum analysis of the noise would help too in distinguish from magnetically induced hum from toroidal trafo [50 Hz], and rectifier's ripple/noise
[100Hz, 150Hz, 200Hz, 250 Hz..]

 

[hotbaby]

This would be an extern pultec power supply with original parts extracted from both mono psu pcbs.


My question is if we can juse just one elco pair and one elco for the 15v section or must we take bigger elcos than?
The picture tells my question again

Anyone?

 

[hotbaby]

ok i just want to say that an extern psu supply seems to be the best solution... we will have no hum problems if we work with smothed dc current in the machine...

 

[keefaz]

Frankly, I would use "scientific" methods to determine where the hum comes from,
With "scientific" way, I mean by experimentation
- Is the hum present when you bypass the filter board, eg: just connect the line amp
- Is the hum present equally on both channels, maybe remove completly one channel and see
- Then I would arrange the AC wires, ground topology, but one step at a time and note the result after each change
I would not jump in conclusion without precise tests...

 

[hotbaby]

hey keefaz, just to make it clear, i tested everything all and came to the point that the rest of hum stays cause of the ac rectiving on the psu pcb in the same case near to the i/o trannys. I could solve all the other hum with grounding and replacing the LED 24v pulsating dc cables which i have disconnect and put them out of the chassi. They humed to the filter high section potentiometers which stayed near to it.

what i tryed else was disconnecting the the left psu pcb and their 250v and heater ac cables and the hum was much lower cause this parallel cable produced a magnetic field over the first psu pcb that had the most hum. All the same with the other ac cables. if i disconnected the cables and lie it out of the case the hum was much more quieter to -74db. It is just logical to me now. Everybody i talked to was saying no ac in audio case. Thats the reason some gurus use Batterys for getting a smooth dc current and thats the point i figured out is the problem with the pcbs if they exist in parallel setup.

I Replaced the toroid out of the chassi but a minimum hum stays. I will try to rectivie out of the case and i am sure i can solve it this way.
All other things like disconnecting the filter, playing around with all grounds i have tryed without useful results. The other thing i point to is the elco job... if the elcos encharges to fast cause of the blind resistors i soldered in there will be a similar sounding power tranny hum too... this could be measured with an good oscilloscope if the dc current after the elcos is flat but i dont have one  :-\

 

[radiance]

ok i just want to say that an extern psu supply seems to be the best solution... we will have no hum problems if we work with smothed dc current in the machine...

So you tried the external PSU then??

Another thing to consider is the fact that there is a lot of unshielded wiring going from and to the Lorlin switches. I see that you've used shielded cable for your audio connections but those  wires to and from the Lorlin are in fact also audio connections. This makes this design susceptible for picking up hum from nearby toroids and such....
Also, why do you need that bleed resistor for? Your "power on" lamp will drain your caps in no time, well at least mine does...

I'm with keefaz as in that you should use this opportunity to really find out what causes your hum using "scientific" methods. I can tell you there is nothing more rewarding than successful troubleshooting ;-)

 

[hotbaby]

>> So you tried the external PSU then??

I found out keeping the ac wires as faar away from the filter potis and i/o transformer will lower this hum which is interfere with the audio signal. The biggest distance with one psu pcb from the ac input to the first sensible parts like the i/o transformer are 10cm...
If i disconnect the right channel / all power cables.in Mono Use of the left channel i have a noise ground which i could live with around -76db (the channel with a bigger space to the powertranny). Yes Mu Metal will do a great job for the right channel too, for getting around -76db i think (putting it around the toroid) but there will stay the ac cable which is above the first psu pcb and on both psu inputs the ac cable will have a distance of just 10cm to the i/o tranny. That is my gole thinking about an extern psu design cause i found no other solution getting the ac hum more far away from the sensible parts. as i have xs1100 which has no mu metal shield this is more sensible than the lundahl one of Dagoose.

The only point i have to figure out now is if the rest of the hum at -76 db comes from a bad "Direct Current" perhaps causes of the resistors or from that, that the rectivers are on the pcbs and to near to the i/o transformers. If 1 i have to displace the resistors or if this doesnt help try bigger elcos. if the second i need to do the same like Dagoose and put the ac/dc conversation on the toroid for a bigger distance and shield it with mu-metall. As Pulsating DC is doing noise too it would be better to put the elcos behind this shield too for working only with smoothed dc in the machine but this will be very tight!

@ radiance like you say, I will figure out every step and try to find which one of these is the evil-doer!!
But please help me for the external ac/dc test, i tryed to explain a question above with a picture if i can just use one elco pair with one 3k3 Resistor for both psu pcbs. i would break out the smoothed dc cable to both PsuPcb points. This could make it possible staying in one chassi without an extern psu case. but 3 elcos more would need to much space...

my powerlamp goes instanly off if i switch the powerswitch off. it is connected to seperate 24v tranny secondarys.

aaron

 

[radiance]

Aaron, I'm not familiar  with the PSU you're using but looking at your drawing I think you can get away with 1 elco pair, provided you split the power after the Elco's off corse ...

 

[hotbaby]

this is the powersupply pcb: http://purusha.smokinggunrecording.com/Dual-Pultec/SRPP.pdf

At moment for stereo use i have two powersupplys inside for each channel... that are 4 powersupplys in one machine! two for ac/dc 250v and two for ac/dc 15v heater... I want to reduce, separate and shield it and see what hum will go away. the elcos are two 220uf/ 250volt each channel paralleled with an 3k3 2w resistor...
Do i need bigger elcos or should it work?

 

[hotbaby]

im 2 minutes away from Netherlands (Herongen -near to Venlo)! be good to me  ;D 

 

[radiance]

First get that toroid out of your box, see if it helps.
Then, you could bypass the PSU on each board you have an make a new one to power both boards.
Those 220uF elco's should be fine I guess.
However, personally I would also try to build another PSU, more like the one found in the DAO project...

http://www.groupdiy.com/index.php?topic=26585.0

And make a small case for it so it can be far away from  the Pultec itself.
This Pultec EQ is a very nice EQ. Why put it in a small box together with a noisy PSU?

Oh, I just moved to the south of France BTW  ;D  Tried to change my profile but it did not work for some reason....

 

[hotbaby]

i can remember when i was a child we stayed for holliday in south France... nice point and the oysters are tasty
The powersupply from volker you pointed to is a really good design. no need for mu-metall as the elco chain seems to shield anough...

As much i would prever a extern design i will try the fast and cheaper thing first using the parts i already have.
I am still shure since i solved serious all magnetic problems that the power supply ac/dc section does the hum. It sounds like that and
the theory of this powersupply cant really work without hum for parallel use.
i will figure out this...