BBD (TDA1022) - Tetrode Gate - correct voltage?

[Thomeeque]

 Hello guys!

According the TDA1022 datasheet this voltage should be set 1 Volt above clock input(s) LOW voltage:



But e.g. R.A. Penfold's Flanger sets there 0.82V (relatively to the ground of the circuit, so it cannot fulfill this recommendation). And Jürgen Haible does not seem to fulfill this recommendation either in his Triple Chorus (but I may be easily wrong here, that's just how I understand the schematic).

So, what is this voltage about, why is not recommended value used and does it worth to try to tweak it (and how? to the recommended value?)?

I'm trying to design and build Electric Mistress based on TDA1022 and I'd like to make it as good as possible of course , but these details are already above my (very basic) knowledge about BBD technology..

Thanks for any entry!

T.

[oldschoolanalog]

In the "Characteristics @" section of the spec sheet the "1V above Clock Low" numbers are based on a supply of -15V.
1V above Clock Low (-15V) = -14V     Clock High being ~0V (Ground)
In the JH design; it appears he is using a diode & a resistor to achieve this V drop at pin 13. Also note he is powering the BBD from a -V power supply. Pin 9 = Vdd; Pin 16 = Ground
Now, in the Penfold design, Pin 9 = Ground; Pin 16 = Vdd.
In this setup (Vdd=+9V), Clock High = +9V; Clock Low = Ground
I don"t know how the 0.82V was derived, but it is above Clock Low.
Lets say this unit was being powered @ +15V. In this case, Clock High would = +15V. Clock Low would = Ground. And, "Tetrode Gate V" would  = +1V.
So, to answer the original Q, yes that .82V value would appear to be correct. Or "In the Ballpark", so to speak.
It would be nice if one of the Real BBD savvy guys would chime in here... (You know who you are )
Hope I didn't create more confusion.
All Things Good,
Dave

[puretube]

Clocks want to be rail-to-rail;

Vg on pin13 = 1/15th of Vdd;
which is +1V,
when Vdd on pin16 is +15V (positive) vs. ground at pin9 (0V).

Derive that 1V through a resistive divider (more exact),
or with a diode-drop ( ~ ).

For negative PS ("positive ground"),
Vg on pin13 = 14/15th of Vdd;
which is -14V,
when Vdd on pin9 is -15V (negative) vs. ground at pin16 (0V).

[oldschoolanalog]

It would be nice if one of the Real BBD savvy guys would chime in here... (You know who you are )

Thanks Ton!

[Thomeeque]

Clocks want to be rail-to-rail;
Vg on pin13 = 1/15th of Vdd

So, do I understand it correctly that the recommendation is not 1 Volt but 1/15th of V_DD above clock input(s) LOW voltage? But this would make 0.6V for V_DD=9V. R.A. Penfold could easily choose either 1V or 0.6V, but he has chosen 0.82V (1/11 of V_DD), that's what I don't get.. Or it simply does not matter so much (drop on diode will be around 0.7V and V_DD=(-)15V in Triple Chorus, so it's 30% away from the recommended value)?

[oldschoolanalog]

So, do I understand it correctly that the recommendation is not 1 Volt but 1/15th of V_DD above clock input(s) LOW voltage?

Yes. Trust puretube's advice. The man knows more about this sort of thing than should be allowed by law.

But this would make 0.6V for V_DD=9V. R.A. Penfold could easily choose either 1V or 0.6V, but he has chosen 0.82V (1/11 of V_DD), that's what I don't get.. Or it simply does not matter so much (drop on diode will be around 0.7V and V_DD=(-)15V in Triple Chorus, so it's 30% away from the recommended value)?

Time to experiment .
I wonder if this is the sort of thing you could hear, or is it only noticable on an o'scope .
Hmm...

[Thomeeque]

Time to experiment .

Unfortunately I have only one piece of TDA1022 (and no easy way to get another one), so I don't want to mess with it too much.. But I will maybe take courage lately, fortunately I have an o'scope.

Clocks want to be rail-to-rail;

Please, could you explain this sentence for me? I've tried to "ungoogle" it, but didn't get very far Or just point me to some explanation somewhere on the web..?

[puretube]

Rail-to-rail =  -15V  to  0V in your case... (powersupply "rails" with negative supply)
or: 0V to +15V for positive supply

or: 0V to 9V for a 9V supply

it just means that the clock-amplitude is as high as the full supply-voltage between pins 9 and 16...


[if you ruin the chip because of a mistake from my side, you`ll have a new one in Praha the day after  ]

[Thomeeque]

Rail-to-rail =  -15V  to  0V in your case... (powersupply "rails" with negative supply)
or: 0V to +15V for positive supply

or: 0V to 9V for a 9V supply

it just means that the clock-amplitude is as high as the full supply-voltage between pins 9 and 16...

Oh, OK - got it, thanks!

But on the other hand, it does not explain this "Tetrode Gate" thing and it's role in the proccess, at least to me

Btw. I did found another recommendation to set it 1 Volt away from V_DD, in SAD1024 datasheet this time (they call it V_BB there):



Again, it's just recommendation without any further explanation what exactly it will do for performance, so I don't even know what should I look for when trying to adjust this voltage (will it optimize bandwidth, dynamic range, distortion, S/N..?)

[if you ruin the chip because of a mistake from my side, you`ll have a new one in Praha the day after  ]

Fair enough!

[StephenGiles]

http://www.4shared.com/file/37718005/8f697e70/TDA1022_c.html
This was given to me many years ago by a Philips designer, it goes beyond the data sheet for the TDA 1022.

[Thomeeque]

http://www.4shared.com/file/37718005/8f697e70/TDA1022_c.html
This was given to me many years ago by a Philips designer, it goes beyond the data sheet for the TDA 1022.

Great! Thanks a lot!

Btw. it's the first time I see TDA1022 based circuit using current source instead of load resistor (datasheet recommends this to reduce delay line attenuation to 2.5dB*, but no schemo there)!

*It still sucks considering fact, that other BBDs don't attenuate at all or even give some boost to the signal, but at least something..

[oldschoolanalog]

*It still sucks considering fact, that other BBDs don't attenuate at all or even give some boost to the signal, but at least something..

Considering that the TDA1022 was developed in 1969 it is truly a relic in the BBD hierarchy. A Great Grandfather of sorts. IMHO, it's performance characteristics are something that one just has to accept and work around. Your idea of adding some gain to the BBD output buffer seems to be a nice simple solution to this issue.
Just my 2 cents.
From a 1 cent kind of guy...
Dave

[StephenGiles]

The additional circuitry offered in the Philips circuit I linked above does in fact go some way to improving the signal to noise factor, so that when the squire makes eyes at the ploughman's daughter - all is not lost!

[Thomeeque]

*It still sucks considering fact, that other BBDs don't attenuate at all or even give some boost to the signal, but at least something..

Considering that the TDA1022 was developed in 1969 it is truly a relic in the BBD hierarchy. A Great Grandfather of sorts.

Oh, that explains few things

IMHO, it's performance characteristics are something that one just has to accept and work around. Your idea of adding some gain to the BBD output buffer seems to be a nice simple solution to this issue.
Just my 2 cents.
From a 1 cent kind of guy...
Dave

Thanks, I'll probably do that - I just must be careful here to not add too much noise to signal since it goes back with feedback and gets boosted a lot.

The additional circuitry offered in the Philips circuit I linked above does in fact go some way to improving the signal to noise factor, so that when the squire makes eyes at the ploughman's daughter - all is not lost!

Yes, I will try to combine this current source trick with small boost buffer, it's probably worthy, when they use it in this circuit.

Btw. would somebody have some tip for this "buffer"? (is it still ok to call it buffer, when it boosts?):

• non-inverting
• gain approx. 1.5 to 2 (+3.5 to +6dB)
• low noise
• input impedance not so critical (>150kOhms?)
• output impedance not so critical (up to few kOhms max.)
• +9(10)V supply

Is there something better/simplier than non-inverting opamp amplifier? Or how should be non-inverting opamp amplifier configured to be as low-noise as possible? Thanks a lot!

Just reminder, all these questions are primarily related to my Electric Mistress based on TDA1022 experiment..

[snap]

+/- flip

[puretube]

Hey, Thomeeque: any damage yet?   

[Thomeeque]

Hey, Thomeeque: any damage yet?   

Not yet, only on my nerves

Please, check the latest status here - as usually, tips and comments are welcome!

Thanks, T.