Helios type69 Veroboard vs Schematic. Narrow "Q" problem.

[Ted Krotkiewski]

The most likely reason for not enough boost is a wrong value pot or 5K1 resistor. These form the basic 14dB pad that defines the maximum boost.

Thank you so much for your help Ian! Cant thank you enough.

Quick report of my further findings.
Firstly, Indeed stupid me, forgot the 5k1 resistor in my breadboard.
The boost is now acting as should. (Yayy)  ;D

Regarding the non-linearity on FLAT setting:
It appears that it is the Bass section affecting the circuit. When I disconnect the Bass section, I do get a "Flat as a pancake" measurement. Even though it is set to zero boost, I get a veery wide (10Hz - 2Khz ! ) boost of about 3db.
Do you have aaany idea (again) what I might do wrong?
(Please see pictures)

Then I also discovered that the general boost specifically the MID was greatly affected with and without Opamp at EQ output. 
With Opamp +15dB
Without Opamp: +19dB
Is this normal? 

Thank you for your time!
//Ted

 

[Ted Krotkiewski]

When putting in the Bass section it becomes like this...  :-\

 

[dogears]

That is not abnormal. You can recreate that in the sim - depending on the value of the gain pot you use in series with the inductor. Later Helios desks had an off switch built into the pot for that reason.

 

[Ted Krotkiewski]

That is not abnormal. You can recreate that in the sim - depending on the value of the gain pot you use in series with the inductor. Later Helios desks had an off switch built into the pot for that reason.

Hi Dogears.
Thank you for your reply.
I tried putting a 500K pot instead of the 50K pot. and the unwanted (10hz - 5k-ish) boost went down to 0.5dBs wich is a great improvement,  8) although not Linear flat as Ian said it should be.

Can´t stop wondering though, what on earth makes Ian´s Freq response flat (with 50K-pot) and mine having this bump?

I am also currently reading up on how to determine "Q" factor value. But by comparing my Width to Ian´s Measurements. My seems to be quite a bit wider? Especially in bass band. Please see my attached very clumsy pics.

Except for recalculating and changing values for LC, is there something else that can be done to a circuit like this to make Q-width more narrow?

This is so much fun, and I am sorry for asking all this basic questions, but for me, this is such a success, making this circuit (almost) work from Ian´s circuit layout.
Can´t thank you peoples guidance enough! 

 

[Ted Krotkiewski]

..And here´s a comparison between my measurement and Ian´s for 6K boost.

 

[dogears]

It won’t be ruler flat with the inductor in the circuit.

 

[ruffrecords]

If you are using the usual switch to select bass boost/cut frequency then you should have a centre 'flat' position where the bass boost pot is disconnected. Once a bass boost frequency is selected there is a minimum  boost determined by the value of the pot. As has been mentions, in some versions of the Helios EQ, the bass gain pot was one with an integral switch to make it easier to turn it off.

If the OP amp after the EQ is affecting the amount of boost then its input impedance is probably wrong. If you are using the TL072 design shown in the Helios documentation on my web site them make sure the input resistor at the + input is 470K.

Cheers

Ian

 

[Ted Krotkiewski]

Hi.
Thank you everyone for all your wonderfull help.
My little EQ build is going really well and I am very happy how it comes out.
It sounds damn beautifull.

I have something I want to try, a little off topic and perhaps a "dumb" question, but thought I´d ask here before spending a lot of time finding it out the hard way myself.

I have a bunch of old Mullard/Philips Mustard caps lying around, that I always wanted to put on a switch together with other cap types to compare if there is an audible difference.

So, I thought that I could create a mirrored row of Mustards (with same values as my current WIMA caps in this circuit).
Lets say I create a duplicate row of all the caps that goes to 0.2H, on the mid band and I put a switch that can toggle between having the current go through WIMA´s or the Mustards.

So my question is, If I for example have the switch set to the WIMA´s, will then the xtra Mustard caps affect the circuit, even though they are switched off/have an open circuit (not leading any current through them)?

I am guessing NO since no current will flow through them and this is how the rest of the circuit switching works, but still. I know nothing, and expect the worst of my ideas, so I have to ask.

Hope what I am saying makes sense.

Having it this way, one could get an honest opinion from myself and other producer friends using it in real life instead of listening to A/B blind tests.
So I thought this would be the best way to do it.

Thank you so much for your time!! 
//Ted

 

[ruffrecords]

If the WIMAs have one end disconnected they will have no effect. It will be interesting to hear if you can detect any difference in the sound. I am not an advocate of using 'vintage' components of any sort and I suspect the sound of the Helios owes more to its design and the inductors than it does to the make of capacitor used.

Cheers

Ian

 

[Ted Krotkiewski]

Hi again.
So, after a lot of listening and experimenting, I have successfully built a prototype that I like very much.
I have changed the frequencies on places to my liking, most significantly up in the high mid & treble area and also in the Low Bass.
I separated the LO-CUT and BASS so they are on separate Switches for more control.
However, since they overlap in frequency it is not very usable wich is why I am writing this post:

Before continuing, I just want to point out that I can imagine that my ever-going questions about this circuit perhaps are getting tiresome, If they are, please let me know and I will respectfully refrain from posting more.

So, to my question.
When looking at the schematics for the original Type69 (V.3) (please see attached pic) There is a LoCut filter applied before the EQ wich is followed by some sort of Gain circuit. I am guessing it is a Unity Gain Buffer?

So, I immediately thought I was smart, and thought, that this must be the reason to why it works in the original where the EQ wont be affected by the Locut.
I am guessing that having separate Gain amplifiers for each band, makes them independent of each other and therefore can overlap. Is my guessing totally off?

So I tried to put up a simple Unity Gain Buffer with a TL072 after the a locut filter, but it does not work. As a matter of fact, when putting a Unity gain buffer after the Locut filter, it does not even "see" the locut change, and gives me a linear output from 10Hz and up. I must be doing something wrong..

However, my question is if my guessing on the subject is somewhat right and if it can be done to this circuit with an Opamp?

On a 1073EQ for example, I love to Locut and Boost lows at the same time to get a "Tighter" low end. I would love to be able to do the same for this design.

Ping Mr @Ruffrecords  8)

Thank you for your time.
My best.
//Ted Krotkiewski

 

[ruffrecords]

HI Ted,

There are quite a few variants of the Helios 69 design. Some have the additional HPF, some don't , some feed the HPF straight into the EQ, later ones have the unity gain buffer. These are all typical evolutionary design changes. There are probably even more variants if you look at the custom desks like the one they made for The Who's Ramport Studios (this does not have the extra HPF).

Using EQ is very much a matter of taste and it is well known that the Helios EQ was liked by musicians because they thought it was more musical. Adding a buffer between stages of an EQ will isolate the responses so there will be no interaction between controls separated in this way except where their responses coincide - it will not stop a bass boost interacting with an isolated bass cut. However, interaction is sometimes a good thing. The Pultec EQP1A for instance has separate bass boost and bass cut controls and a common frequency knob. Even though the boost and cut frequencies are labelled identically, in practice their turnover frequencies are different. This means the boost and cut do not cancel each other out but produce the well known "Pultec Bump" where the response dips at a higher bass frequency and is then booted at a lower one.

One thing to bear in mind is the additional HPF has two cascaded sections so it cuts at 12dB/octave rather than the gentler 6dB/octave of the low cut in the main EQ. It is really intended for removing unwanted low frequency noise and rumble than as an EQ tool. But there is no reason you cannot add an op amp directly after the low cut control in the main EQ. You will need to retain the 180K resistor to ground and add another one after the op amp.

Cheers

Ian

 

[Ted Krotkiewski]

Hi Ian.
Thank you!! Very informative read.

I am not sure I follow, Perhaps due to my English.
I will trouble you with a follow up:

- it will not stop a bass boost interacting with an isolated bass cut.

So, if adding a Unity Gain buffer after the LoCut filter does NOT make them Independent from each other,
Is there a simple way to make the EQ circuit "not react" to the Lo Cut section before it? So that when boosting the bass, it simply adds bass at given frequency "without knowing" that it has been Lo´cutted.

Adding a buffer between stages of an EQ will isolate the responses so there will be no interaction between controls separated in this way except where their responses coincide

It makes me wonder, If two bands are overlapping and still behaves in the same way with or without a buffer.
What is then achieved by adding buffer for each band?
If I am not mistaken the Siemens W295 EQ´s and the 1073/1081 for example has that?

Hope I am not asking for too much at the same time. In my head it seemed so obvious that it would work.
My best.
//Ted

 

[ruffrecords]

OK, I think I need to explain it in general terms. A passive like the Helios EQ works by making an attenuator for the incoming signal. In the Helios it is basically the 22K pot and the 5K1 resistor to ground. This attenuator has a loss of about 15dB.

Boost and cut work in different ways in this type of EQ. Boost works by adding some or all of a frequency selective network in parallel with the 22K. At some frequency the network might be  nearly a complete short circuit so that instead of 15dB attenuation at the frequency, you get 0dB i.e. you have boosted the signal by 15dB.

If you have two boost circuits at similar frequencies so their responses overlap such that they for example each produce 10dB of boost at the overlap frequency, you will not get 10dB + 10db = 20dB of boost at that frequency because the best you can do is short out the 22K and get 15dB of boost. So these two boosts do interact. However, if each one had its own 15dB attenuator then they could each produce 10dB boost at the common frequency and hence 20dB overall.

Cut is different. You can cut by adding a frequency  selective network across the 5K1 resistor. If your network has an impedance of 1K at the cut frequency you get the total attenuation increase from 15dB to about 27dB ( ~1/23) which is an attenuation of 12dB. If you have another frequency selective circuit that also has a 1K impedance at this frequency the total attenuation becomes ~0.5/20.5 = 32dB. So although the two 12dB attenuations have not added, the total has increased by 6dB.

So to summarize, cuts and boosts using frequency selective networks across the 22K or the 5K1 resistors will interact with each other.

However, the Helios bass cut is done differently. It is a simple CR circuit at the input to the EQ. It does not rely on a frequency selective network across either the 22K or 5K1 resistors but it does assume its load is 22K +5K1 when it does its cutting. And because the bass boost cannot operate while the bass cut is operating, its load is never less than this at the frequencies it is cutting.

Lastly we come to the external HPF. Because the bass boost could be operating at the similar frequencies to the HPF, you can no longer guarantee the HPF load is 22K + 5K1, so you do need a unity gain buffer. This means the HPF will be unaffected by the bass boost or cut and the total boost or cut will simply be the sum of the boost/cut in each stage.

Bottom line is the frequencies in the original H69 EQ are carefully chosen to minimise any interactions between the bands - in other words they avoid frequency overlaps. The add-on HPF is the exception and needs to be buffered to prevent unwanted interactions.

Cheers

ian

 

[Ted Krotkiewski]

Woaaaow!!
Amazing that you took your time writing this down for me.
I will propably head back here and re-read what you just wrote for many years ahead.
This is Golden information!

Lastly we come to the external HPF. Because the bass boost could be operating at the similar frequencies to the HPF, you can no longer guarantee the HPF load is 22K + 5K1, so you do need a unity gain buffer. This means the HPF will be unaffected by the bass boost or cut and the total boost or cut will simply be the sum of the boost/cut in each stage.

Bottom line is the frequencies in the original H69 EQ are carefully chosen to minimise any interactions between the bands - in other words they avoid frequency overlaps. The add-on HPF is the exception and needs to be buffered to prevent unwanted interactions.

So, if am I understanding this right, the added HPF (With a buffer afterwards) does Not affect the EQ circuit afterwards only because it is BEFORE the "22k + 5k1" EQ network?
So, in other words, If it´s set up in this way, can I then achieve a boost at lets say 30Hz even though I have the added HPF engaged?
Next week I will be able to test this on my breadboard. In the meantime, Im asking it here.

Forever thankfull!
//Ted

 

[Rocinante]

OK, I think I need to explain it in general terms. A passive like the Helios EQ works by making an attenuator for the incoming signal. In the Helios it is basically the 22K pot and the 5K1 resistor to ground. This attenuator has a loss of about 15dB.

Boot and cut work in different ways in the type of EQ. Boost works by adding some or all of a frequency selective network in parallel with the EQ. At some frequency the network might be  nearly a complete short circuit so that instead of 15dB attenuation at the frequency, you get 0dB i.e. you have boosted the signal by 15dB.

If you have two boost circuits at similar frequencies so their responses overlap such that they for example each produce 10dB of boost at the overlap frequency, you will not get 10dB + 10db = 20dB of boost at that frequency because the best you can do is short out the 22K and get 15dB of boost. So these two boosts do interact. However, if each one had its own 15dB attenuator then they could each produce 10dB boost at the common frequency and hence 20dB overall.

Cut is different. You can cut by adding a frequency  selective network across the 5K1 resistor. If your network has an impedance of 1K at the cut frequency you get the total attenuation increase from 15dB to about 27dB ( ~1/23) which is an attenuation of 12dB. If you have another frequency selective circuit that also has a 1K impedance at this frequency the total attenuation becomes ~0.5/20.5 = 32dB. So although the two 12dB attenuations have not added, the total has increased by 6dB.

So to summarize, cuts and boosts using frequency selective networks across the 22K or the 5K1 resistors will interact with each other.

However, the Helios bass cut is done differently. It is a simple CR circuit at the input to the EQ. It does not rely on a frequency selective network across either the 22K or 5K1 resistors but it does assume its load it 22K +5K1 when it does its cutting. And because the bass boost cannot operate while the bass cut is operating, its load is never less than this at the frequencies it is cutting.

Lastly we come to the external HPF. Because the bass boost could be operating at the similar frequencies to the HPF, you can no longer guarantee the HPF load is 22K + 5K1, so you do need a unity gain buffer. This means the HPF will be unaffected by the bass boost or cut and the total boost or cut will simply be the sum of the boost/cut in each stage.

Bottom line is the frequencies in the original H69 EQ are carefully chosen to minimise any interactions between the bands - in other words they avoid frequency overlaps. The add-on HPF is the exception and needs to be buffered to prevent unwanted interactions.

Cheers

ian

Ian
This ir truly a great summery of the Helios EQ circuit. Thank you.

 

[ruffrecords]

Woaaaow!!
Amazing that you took your time writing this down for me.
I will propably head back here and re-read what you just wrote for many years ahead.
This is Golden information!
So, if am I understanding this right, the added HPF (With a buffer afterwards) does Not affect the EQ circuit afterwards only because it is BEFORE the "22k + 5k1" EQ network?
So, in other words, If it´s set up in this way, can I then achieve a boost at lets say 30Hz even though I have the added HPF engaged?
Next week I will be able to test this on my breadboard. In the meantime, Im asking it here.

Forever thankfull!
//Ted

That is right. On re-reading my post I have made a couple a small tweaks to make it hopefully even clearer.

Cheers

Ian

 

[Ted Krotkiewski]

That is right. On re-reading my post I have made a couple a small tweaks to make it hopefully even clearer.

Wonderfull, Ian.
I finally understand how it all works. (Kind of..)
In order to isolate one band with it´s own buffer, one needs to create it´s own 22K + 5k1 network! (except for the Locut ofc.)
I immediately tried to isolate the High Shelf due to the overlapping occurring with my my added high-Mid freq´s (8.5k,11k,16k).  ​
So, with it´s own network I succeeded making the High Shelf work as a single High shelf "Plugin" 
Bloody awesome!

However, For some weird reason, whenever I insert a Unity Gain buffer It strangely "resets" whatever has been done before it and the freq response gets completely flat. So, if I have -15dB of Locut, as soon as I add the Unity Gain buffer after the Capacitor, it gets flat. It does not matter whether I do it stand alone, or in line with the EQ circuit.
But this lends itself to another Op Amp help thread me thinks, as it has nothing to do with the 69 circuit.

This is so much fun.
Btw, here´s a pic of my prototype.

 

[ruffrecords]

The low cut consists of a series cap followed by a resistor to ground which in the Helios 69 EQ is the 22K in series with the 5K1. So when you insert the unity gain buffer you need replicate this resistor else the cut will not work.

Cheers

Ian

 

[Ted Krotkiewski]

The low cut consists of a series cap followed by a resistor to ground which in the Helios 69 EQ is the 22K in series with the 5K1. So when you insert the unity gain buffer you need replicate this resistor else the cut will not work.

Thank you Ian.
Apologies, but I am totally lost here.
In what way should I add/replicate the 22k + 5k1? and where?
I thought that the Lo-Cut was independent of the 22k + 5k1 network, since nothing goes parallel with the filter?

The thing is, I get the same result even if I just set up a signal loop through a simple 22n cap filter back into the soundcard.
Without Unity buffer. It filters the signal. With Unity gain, it becomes flat.
So the same thing happens with or without the 69 Circuit.

Same thing happens if I put an additional High Shelf band with it´s own 22k + 5k1 network AFTER the EQ output, It "Resets" any EQ adjustents that has been done before the buffer.
hmmm.. Im lost.

Thank you for your time.
My best.
//Ted

 

[ruffrecords]

As I said, the low cut consists of two components, the series capacitor and a resistor from the cap to 0V. I have attached a pic of the schematic for a simple low cut. I have used a single 27K instead of 22K + 5K1. Connect this from and to your soundcard and see what you get.

Cheers

Ian