stage 1.5 schematic+pcb amek2500eq+harrison filters

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

12afael

Well-known member
Joined
Aug 6, 2004
Messages
1,330
Location
Helsinki
inspired by analag´s post (this beautiful forum :roll: ) I put myself to work in a mix of two project that were posted time ago.

http://www.groupdiy.com/index.php?topic=20972&highlight=amek+2500+filters
and
http://www.groupdiy.com/index.php?topic=4455&highlight=amek


AMEK2500RENDER.jpg


Features:

- Balanced input and output using INA134, DRV134, SSM2141, SSM2142 , That 1250, That1646 or equivalents.

- 4 band full parametric amek 2500 EQ:
- each band has Q, frequency and gain controls.
- Shelf/bell option for HF and LF bands(optional).
- NEMEK option, use dual linear pots instead the dual rev logs!
- Eq bypass switch(optional).

- Harrison filters(optional):
- Second order low pass filter. (-3db at 170 Hz to 19kHz)
- Second order high pass filter. (-3db at 24Hz to 3.3kHz)
- Filters bypass switch (optional).


I don´t have a lot of experience designing so I want to hear opinions of the schematic. Mainly about impendances and stage coupling between the circuits of the amek, harrison and IN/OUT.

at this moment the project is:
- A single layer pcb of 30x10cm (maybe 2 boards of 15x10), harrison filter pots and switches out of the pcb.

- 5 pages documentation file(options, description and partlist).

normal_schematicamek2500.gif


Click here for full size.
http://www.twin-x.com/groupdiy/albums/userpics/schematicamek2500.gif

The pcb board is 90% finished but I want to be sure of the schematic to release it.

greetings

Rafael
 
I looked at it some and didn't see the bell/shelf option in the schematic.

Also my preference with adjustable state variable sections is not to use the frequency potentiometers as variable resistors (20%) when you can use them more precisely to establish ratios.


JR
 
John, thanks for the post, more than 100 views without a post...

JP3 and JP4 are switch to bypass the 220K Rs, with the 220K the Q control don´t do anything, lsp16 and lsp17 are pads to solder the bell/shelf switch . this is an optional part so you must cut the trace between the pads.
lsp14 and lsp15 are pads for the LF bell shelf switch, they bypass the 1M R, if you don´t want the shelf bell put a jumper on R53.

Also my preference with adjustable state variable sections is not to use the frequency potentiometers as variable resistors (20%) when you can use them more precisely to establish ratios.

that´s the NEMEK (NEve aMek)option (cut a track and put 2 jumpers). use a dual linear pot to make a voltage divisor.
there is a difference on noise performance?

I was thinking that c23 and R44 should be part of the output circuit. if the circuit is bypassed there is not decoupling between the INA134 and drv134. I don´t know if offset is a problem here(reading datasheets :roll: ).

should I ground the input if a circuit is bypased?

what value of R44 would be reasonable? the input impedance of the DRV134 is 10K

Rafael
 
I'm not sure if the bell/shelf is fully implemented... the second half of a DPDT switch should be used to bypass (short) a 1-meg in series with one half of the frequency pot for one shelf, and the other shelf should have one half of the dual pot completely bypassed...

I'm not CERTAIN that I've figured out how to make the NeMek (potential divider) wiring method work... you may need a third pole of the switch for one shelf...

Still thinking about that one...

Keith
 
[quote author="12afael"]John, thanks for the post, more than 100 views without a post...

JP3 and JP4 are switch to bypass the 220K Rs, with the 220K the Q control don´t do anything, lsp16 and lsp17 are pads to solder the bell/shelf switch . this is an optional part so you must cut the trace between the pads.
lsp14 and lsp15 are pads for the LF bell shelf switch, they bypass the 1M R, if you don´t want the shelf bell put a jumper on R53.

Also my preference with adjustable state variable sections is not to use the frequency potentiometers as variable resistors (20%) when you can use them more precisely to establish ratios.

that´s the NEMEK (NEve aMek)option (cut a track and put 2 jumpers). use a dual linear pot to make a voltage divisor.
there is a difference on noise performance?

I was thinking that c23 and R44 should be part of the output circuit. if the circuit is bypassed there is not decoupling between the INA134 and drv134. I don´t know if offset is a problem here(reading datasheets :roll: ).

should I ground the input if a circuit is bypased?

what value of R44 would be reasonable? the input impedance of the DRV134 is 10K

Rafael[/quote]

OK I follow now, the shelf is still using the BP output from the state variable just pushing out the LP or HP pole on end sections. When I put parametric shelf in a console back in late '70s, I actually grabbed the appropriate HP or LP output of the SVF. They are opposite polarity from BP so other changes were involved too. Your way has merit if you want to make the shelf EQ go away out of band, I usually try to just make the boost go away above/below band. I also damped the Q since the concept of adjustable bandwidth for a shelving filter is harder to make sense of and use.

I am not familiar with that output driver so have no suggestions wrt that.

JR
 
I'm not sure if the bell/shelf is fully implemented... the second half of a DPDT switch should be used to bypass (short) a 1-meg in series with one half of the frequency pot for one shelf, and the other shelf should have one half of the dual pot completely bypassed...

is not so clear on the schematic because it is an option and you need to cut a trace to implement the shelf/bell.

I'm not CERTAIN that I've figured out how to make the NeMek (potential divider) wiring method work... you may need a third pole of the switch for one shelf...

I´ve tested the shelf/bell with the amek schematic on ltspice, I will do the same with the nemek schematic , I will have an answer this afternoon.
 
[quote author="12afael"]- Harrison filters(optional):
- Second order low pass filter. (-3db at 170 Hz to 19kHz)
- Second order high pass filter. (-3db at 24Hz to 3.3kHz)
- Filters bypass switch (optional).


I don´t have a lot of experience designing so I want to hear opinions of the schematic. Mainly about impendances and stage coupling between the circuits of the amek, harrison and IN/OUT.[/quote]
FWIW: in the minium positions of the filter-pots the load on the TL-opamps may be a bit too much (see IC5, as much as 390 Ohms for the relevant frequencies), while a rule of thumb says ~2k for TL07X-stuff.

So while it isn't too healthy from a technical point of view, I do realize that lots of designs do this, so FWIW...

Bye,

Peter
 
In IC5 that <2K load is primarily at highest frequencies so not as much of an issue, but not good if input not adequately bandpassed.

Agreed 2k is preferred minimum loading for bifet series. Perhaps values could be scaled if that much range is really needed. This may be another benefit of potentiometric rather than rheostat topology.

To answer an earlier question "does this make a noise difference"? No. The primary difference is accuracy, precision, and repeatability in production of frequency adjustments. Bulk resistance of pots is typically 20% while ratio accuracy much better.

JR
 
[quote author="JohnRoberts"]In IC5 that <2K load is primarily at highest frequencies so not as much of an issue, but not good if input not adequately bandpassed.

Agreed 2k is preferred minimum loading for bifet series. Perhaps values could be scaled if that much range is really needed. This may be another benefit of potentiometric rather than rheostat topology.
[/quote]
You're right, it's by virtue of the caps that it's always the highest freqs that see this 390 in its full glory, and since these are less likely to have full swing (apart from synthetic sources) it'll likely be fine (more music-stats in the recent RIAA-thread from PStamler).

Using this topology and change caps for coils to change LPF into HPF etc won't be a good idea though, it'll maximize 'the problem'. But there are already other reasons not to use coils of course, so again, it'll do.

Bye,

Peter
 
an opa4134 have an output current of 35mA but the datasheet don´t show a "MAXIMUM PEAK-TO-PEAK OUTPUT VOLTAGE VERSUS LOAD RESISTANCE"
it say
OPA134 series op amps are unity-gain stable and provide excellent
dynamic behavior over a wide range of load conditions,including high load capacitance.
:?

the 4741 op amp was the original used by harrison.

TL074
TL074.gif

V/S
4741
4741.gif


Rafael
 
I scale some values of the harrison filter: 390s to 3.9k, pots from 47k to 470k, c37-38 to 0.01uF , c5 to 4.7n and c4 to 1.2n

the dual rev logs are not easy to find, www.audiomaintenance.com just have have 470K dual rev logs on carbon. I could have changed the values in a factor of 2 to use pots of 100K, but the resistance would even be under 2k.

what other op amp can drive low impedances?

Rafael
 
Keith the shelf/bell switch don´t work on the HF band on the NeMek circuit.

you may need a third pole of the switch for one shelf...
mmm to switch the pot to the old amek circuit? it will not work because the new pot is linear. I must read more about the SVF.
 
[quote author="12afael"]I scale some values of the harrison filter: 390s to 3.9k, pots from 47k to 470k, c37-38 to 0.01uF , c5 to 4.7n and c4 to 1.2n

the dual rev logs are not easy to find, www.audiomaintenance.com just have have 470K dual rev logs on carbon. I could have changed the values in a factor of 2 to use pots of 100K, but the resistance would even be under 2k.

what other op amp can drive low impedances?

Rafael[/quote]
It may not be that essential to avoid the 390, see the comments from John above. But if you can increase it then why not ? You want continuous freq ?

Or how about some discrete frequencies ?
( See for instance the hi- & low-sections of http://www.gyraf.dk/schematics/KH_UE-1000.pdf )
For say twelve frequencies use a DP12T-rotary switch and make any law you want. Such switches that fit 1RU are not hard too find.


Bye,

Peter
 
yeah we all love dual op amps but the use of quads help to do layout simpler, in addition you need less caps to decouple the power supply.

Peter, I will include the discrete option on the documentation,
 
Keith, I re check shelf/Bell switch and it seems to work well. I lost a lot of time with the transfer function of the nemek and finally it is equal to the amek :? .

here is the new schematic, I add the bypass switches (now on pcb!) and I change the position of the output RC net.
http://www.twin-x.com/groupdiy/albums/userpics/amek2500eq%2BharrisonFv2.gif

here is the pcb
http://www.twin-x.com/groupdiy/albums/userpics/nemekpcb.gif

and here is the ground plane highlighted.
http://www.twin-x.com/groupdiy/albums/userpics/nemekGND.gif

so what can I improve?

probably I am going to divide the pcb in two, a 30x10 is hard to etch at home.

Rafael
 
is time for a report. the eq is working but it has some issues.
just for test I`m using two linear 50k dual pots on the harrison filter.
the low pass filter is working ok but the high pass have a big squeal on high freqs, I must check if is only a wiring issue or a design mistake.

for the Eq section I`m using dual linear 50K pots in NeMek mode.
Q controls work well and +-gain too , the frec control make a pop and then no sound comes through the eq if you put the freq control on a low freq.

the shelf bell control on the lower band seems to work ok , but on the high band it make the same pop and no sound of the freq control. I must re check my wiring tomorrow.

I will try to put some pictures and a freq plot soon.

Rafael
 
just as report and because some people is asking about the status of this proyect...

I have solved the problems, the NEMEK(amek2500eq but with dual 50K linear pots instead of rev log50k pots) is working.

Colin (slanderchap, www.audiomaintenance.com) was so kind to send me some dual 50 rev logs to test the stock amek circuit.
unfortunately I haven`t had time nor I could get a 1U rack to test the original option(with rev log pots).

the harrison filter work ok too, but I've only tested with linear potentiometers, I don`t know if with rev logs the rotation/freq feel better, I must finish the second unit (with dual rev log pots)to be sure.

I `ve been working on the pcb too IIRC the current vercion is a two layer pcb.
A pdf with info about the project is being written too.

at this moment I`m a bit busy with a new work and finishing my own album.

I hope in a month or so have some time to continue working on this project if there are people interested.

Rafael
 

Latest posts

Back
Top