NewYorkDave's Passive Equalizer

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
> around 850mH and the other 835mH

In an L-C tank, this gives a 1:1.008 difference of frequency. Say 1/6th of a semi-tone. How can that matter?
 
hi PRR!

i didn't reallies that it was being that precise! thats great i guess. haven't done the math and i don't know how daves eq work entirely... just wanna get it as matched as possible but are we talking better then quarter tones then I'm pretty happy!

the 820mH inductor is used in a shelf filter if I'm not wrong. so wouldn't that be a RL filter rather than a LC...

my measurements are not exact anyway cause with my analogue oscillator and oscilloscope i get a little different value from every sweep...

thanks for your notification!!
 

Attachments

  • 20161023_180147.jpg
    20161023_180147.jpg
    1.9 MB · Views: 48
hi all!

i am trying to calculate the resistor ladders for 1db steps instead of 2db steps.

for the boost ladder, i think I'm right in my calculations but I'm not sure I'm thinking in the right way with the cut ladder.

should the total  resistance in the cut-ladder, be calculated in parallel with the load resistor (600ohm). and then use this new value as the load/shunt resistance in a L-pad?

when i do this and try to confirm with Daves schematic(2db-steps) i get almost the right values... something like 10-30 ohms off..!

i haven't included the DCR of the inductors.
however since they vary, i dont now what to aim for if they should be included...
trying to find info about this but thought i could ask the teachers of this forum :)

attached a picture of where i was a couple of days ago.
 
yesterday, i hooked up one channel with scope, oscillator and dmm, without any makeup gain.

i put a 2k2 load on the output and played with a trim pot to find my desired eq gain points. i feel a bit like a cheater but it feels good to know that the uncommon resistor values i ordered are gonna work and that i won't be needing to desolder them and order a new set of resistors.

there are some physical tones coming from one or several inductors, when feeding the eq with square waves at a certain amplitude.

it seems that the broadness of low shelf looks different depending on how the mid is setup. 40hz boost starts at 600hz or even a little higher, but if i boost the mid band at 400hz, the low shelf starts at 170hz. and so on. very cool!

waiting for mains transformer and resistors for the boost/cut ladders.
 
I'm having thoughts about the grounding of the eq circuit. does it need to be grounded? or does it work in a balanced way.

in the layout with the tube amp that Dave made on page one, the ground is connected with the electronic ground of the tube amp.

but in the initial circuit, when the signal is to be sent to an external unit for gain, with or with out an isolation transformer, there is actually no ground.

i was planning on having the possibility to use either the tube amp efter the eq, or to send the signal elsewhere via relays and separate xlr contacts . if i isolate them, the eq looses the ground plane from the tube amp. would the eq still work when sent to an external balanced mic pre for instance?
 
the eq circuit is done. still waiting for a mains transformer for the tube amp.

i currently have the eq hooked up to a JLM pre amp, with adequate fixed gain and with a carnhill vtb9049 OT for some initial listening.

sounds a bit like heaven.

very interesting eq. i feel that its almost like a one band eq because all the bands interacts with each other.
for ex, if id like to have an impressive low end, id try cutting the 300hz or 400hz and perhaps adjust the lows to taste.
some of the highs and perhaps 1k might need a bit wider Q.

i have pushed it a bit with music, like 14-16dBu, but haven't heard any obvious hard saturation. its actually quite transparent but in the same time euphoric and take cares of the edgy stuff in the mix. perhaps its the carnhill OT that gives that deep and sweet complete image. I'm very pleased with current results of it.

i noticed that i get some low volume hum when the unit sits on my adda and have dialed in some cut/boost in mid and lows. its not loud enough as it is but if theres for ex, lots of compression or silence in the music it could be a problem. so I'm thinking about some shielding work using thick pieces of metal under and over the inductors. perhaps even relocate the inductors so they take place in the middle of the box, to get them as far away from radiation from beneath and above.

however if i remove the eq from the adda the hum goes away, so worst case scenario is to find its place in the rack.

love this eq!
 
i finally got a mains transformer a couple of days ago.

have been assembling the back piece with the psu on two pcbs and the transformer on the outside of the case. its not very beautyfull but I'm quite pleased with it.

i made a fault at first having the 7805 backwards (reversed in and out) and also had some copper flakes, slightly bridging some leads on the breadboard. there was at one time a  buzzing sound for a sec from something...

if i put some load on the regulators i loose dc volts pre and post regulators and also gain in ripple. what could the cause of this be?

the only thing i can think of now is that i got too low filtering pre regulators or that my lm338 got fried. ill see if i can get parts tomorrow.

however, i have a feeling that my transformer might not be up for the task. the winding for heaters(6vdc), power indicator and relays(5vdc) is rated 6.3VAC CT 1.8A.

im not very good at calculating psu requirements.

if i hook up the heater of one tube with AC, direct from the transformer winding, everything else looks much better around the regulators. so in worst case scenario i could have heaters hooked up with ac and try wiring as good as possible. but Id like to have as low noise as possible and with so many inductors and transformers it seems like a good idea to have everything fed with DC.
 
i ripped most of my 6V and 5V circuits and made another run.

heaters run on AC directly from the secondary winding. super tight twisted cables.

relays and power indicator lamp is driven by a single 5VDC line, filtered by 4700uF and a new IC regulator i got from my local store today. the lm2940 5V low drop regulator. all seems fine now.

i should probably got a transformer with a higher voltage from the beginning to make all that DC distribution work as planned...

attached is a schematic of how the psu looks for the moment.
 
the psu and tube amp -  that Dave  designed -  is done!

I've been listening a couple of days to it. the details, enhancement and image is talking over my room. very vivid! after awhile i got the feeling that it was too much. almost felt a bit nauseous and getting a bit tired in my ears.

maybe something was not alright with it.

hooked it up with scope. and yeah, things seemd not right.

heres two 10khz square waves going through the whole unit. bottom wave is after the input transformer and top wave after the interstage transformer before the tubes.
 
and heres how it looks after treating the interstage transformer with some small series resistance on the primary side, loaded the secondary quite a lot and lost some amplitude (but there should be plenty of gain left).

also put a zobel network on the input transformer. the interstage transformer ll7906 i got might not been the best choice...

haven't listend yet...


i am pretty stunned how well everything works besides all this. i can't hear any hum or anything. Plate voltage maybe has 3mv of ripple..! connecting the eq to my adda and short the inputs pin2 and 3, my computer says it detects noise floor around -74dbfs..! and theres additional gain about 6-9db...

can it actually be this good?!?!
 
so the sound of this eq and tube amp is quite transparent. i get a wider soundstage which seems sweet but i think the lo mid and bass couldve been tighter.

i wonder if could improve things with one or two things.

- adding  series resistors to the output transformers (1k or something)from the mosfet regulated B+ line, and perhaps a decoupling capacitor at both centertaps. could this improve on crosstalks, 120hz hum, bass definition?

- making the cathode bias of the amps with constant current sources instead of the bypassed cathode resistor.  what would the profits from this be?

 
I know this is an old thread and there is much respect probably still needed for guys like New York Dave...
At any rate I've been gradually moving my builds towards more of a mastering set up and I have the schematic and about 90% of the parts for this EQ that I am diggin out of the closet so I can finally get around to building this.

My first phase will be to get the kit I got from a member here working with just the Edcor transformers...phase 1
Phase 2 will be to use Ruff Records twin tube kit to be the bookends to the EQ.
I think I can manage most of this.


The EQ kit is a bit of a rats nest since it was always designed to be p2p...so there will be a few days of tracing wires to resistors to switches to inductors and on to transformers/etc...

Here's the question I have going on in the back of my mind:

When I look at the Q chart Mason was so generous to provide in post #41 this thing reminds me of an API 550 proportional Q...although its not as consistant...it seems like its an API 550 on steroids...

Not EXACTLY proportional but its definitely not constant either...the thing that looks really cool, but sort of confuses me is why in the EQ plot does the middle band appear so much narrower than the low/high bands?

Was that intentional? If it was it is brilliant on a different level.
 
I think Ian all of NY Daves files were removed, I have an actual paper print out I did from years ago...its basically a resistor divider going into the inductors via stepped switches...the EQ curves was what I was asking about if you look at the attachments in post #41 they look to be a proportional Q...
 
Turns out I have a copy too. Now what was the original question??

Cheers

Ian
My Question was basically about the EQ curves...I notice that it very much follows the idea of a proportional EQ like the API 550 ...but the slopes are not exactly what I would call "perfectly proportional" (referring to the slopes in post #41)

The midband (according to the pic) is quite a bit steeper...which seems like a good choice...

Is this curve more a product of the inductor being selected with the resistors or just how the bands overlap?

Why is the mid so "pointed" compared to the other two bands?

By the way there is a very (better than watching the boobtube) interesting conversation with Paul Wolff re: API history found here if you guys haven't seen it...T Bone Burnett made a small fortune reselling an API console that he had used to make a lot of money in the first place...here


NYDEQ slope.png
 
Last edited:
Back
Top