TDA1022 Flanger

[Ice-9]

I am posting this Flanger circuit designed by R.A. Penfold, there has been  a few requests here for a circuit that uses the TDA1022 BBD device. I have drawn the schematic in two parts, the first is the signal part and the second is the clock/ modulator part. I built this myself a few years back and it sounded very good. If i was to build it again i might change the clock modulator circuit a  bit. It works fine from  a 9v battery but maybe would be better from a higher voltage.

The main signal schematic



The clock generator

[oldschoolanalog]

Thanks for sharing this.

[Thomeeque]

I am posting this Flanger circuit designed by R.A. Penfold, there has been  a few requests here for a circuit that uses the TDA1022 BBD device.

 Thanks a zillion times, Mick!

 Could you clear for me few points, please, yet?

• I don't see dry signal going to output, is it stand-alone effect as it is or is it wet path only (maybe module for something bigger?)? This would flange (thanks to non-cancelable feedback) I guess, but it's quite unusual design.
• Should not feedback go to pin 2 of IC1a instead of pin 1?
• Should not pin 6 of IC5b lead somewhere?

 Thanks, T.

[Ice-9]

I'll try and answer those questions

1. This is the complete effect (no need for a mix of dry signal) I always draw my schematics like this (my laziness) to get the effect to switch on/off all you need is to put a bypass switch into the circuit on the input/output for true bypass. If you want i can redraw the schematic to include the switch
2. Yes pin 6 oc IC5b should go to the voltage divider (same point as IC5a pin 3)
3. Also yes the feedback should go to pin 2 (not pin1)

I have checked the rest of the schematic and its all correct apart from points 2 and 3 you mentioned.

[Thomeeque]

I'll try and answer those questions

1. This is the complete effect (no need for a mix of dry signal)

 OK, hmmm, it's really first time I see flanger which does not mix dry (real-time) signal with delayed/modulated signal, it must sound bit different, interesting! On the other hand it means you can't get chorus-like sound out of it and even light flanger sound (with very small or none feedback) as well. What I will probably do will be surrounding this circuit by input and output buffer with pre-/de-emphasis and dry/wet signal mixing pot (and full control of feedback amount). It will loose part of it's simplicity (and originality), but will by more classical flanger (it will still allow original setup).

 Btw.  what exactly sets 47k trimmer on 555? Is it base frequency? Than it should be possible to make MANUAL pot out of it, right?

I always draw my schematics like this (my laziness) to get the effect to switch on/off all you need is to put a bypass switch into the circuit on the input/output for true bypass. If you want i can redraw the schematic to include the switch

 No, please, that's really not necessary

2. Yes pin 6 oc IC5b should go to the voltage divider (same point as IC5a pin 3)
3. Also yes the feedback should go to pin 2 (not pin1)

I have checked the rest of the schematic and its all correct apart from points 2 and 3 you mentioned.

 Perfectamundo! Thanks again, Mike!

[Ice-9]

Yes it's a bit unusual to not have the dry signal mixed into the output. You could always add an extra mixer/buffer stage at the end to add in the dry signal. Also put in a spst switch in the feedback path to switch out feedback/regeneration this would switch it between a flanger and a chorus. Also swap the 47k trim that controls the 555 timer for a pot to alter the delay time.

[Thomeeque]

 Great!

 Maybe one point yet - could you explain your statement "If i was to build it again i might change the clock modulator circuit a bit.", please?

 You did not like sweep of the original circuit? It is probably pure triangle (not "hypertriangle"), right? Or is there other reason?
 

[Ice-9]

The clock / modulator was fine from how i remember the effect, the sweep was ok (triangle) I would just like to build the clock using a diff design as i think having to use the 4011 and 555 timer combination used a lot of pcb space. It would be nice to try a hypertriangle.

[Thomeeque]

Just small note, I have come across another flanger not mixing dry signal into the output: YAMAHA FL-10MII

[StephenGiles]

Just small note, I have come across another flanger not mixing dry signal into the output: YAMAHA FL-10MII

But it must be, otherwise it won't flange.

[Thomeeque]

Just small note, I have come across another flanger not mixing dry signal into the output: YAMAHA FL-10MII

But it must be, otherwise it won't flange.

 It will flange thanks to feedback, but it will be definitely different.. btw. you can try it with my digital mistress

 Edit: Yep, it does: FLANGED_0%_DRY_68%_FB.wav 

[Thomeeque]

 Hi, all!

 These days I'm working on TDA1022 based Electric Mistress design - it's basically Tonepad's 9V EM penetrated by the R.A. Penfold's TDA1022 flanger above (from where I have derived "only" circuitry around TDA1022 and the idea to use V_ref as BBD DC BIAS as well).

 Actual schemo looks like this:

 

 I have already "eagled" PCB as well, I'll build it next week. My believe is, that on principle it should work, even some tweaks will be necessary probably. I see these arguable point now:

• I'm not sure yet if same clock frequency generates same delay time on TDA1022 as on SAD1024 - hopefully C14 capacity doubling/halving will be sufficient to fix this..
• Output loudness will by maybe lower compared to true-bypassed signal - either I will try to play with total gain of the circuit (R1/R2 divider will be place to start) or I will get over true-bypass and use same trick as original 9V EM (which seems to quiet down bypassed signal to match loudness in effect and bypass modes) or I will use some external buffer or whatever, there will be some way for sure.
• V_ref used as BBD DC BIAS as well can lead to lost of some headroom if DC BIAS will be too far from 4.5 volts - in this case I will rebuild it to get V_ref and BBD DC BIAS independent (as in original 9V EM).
• I have added flasing dual color LED speed indicator, but it should be harmless - I did already sucesfully used it before on my Phase 90, works nice and does not audibly influence sweep.

 Comments are welcome!

 T.

[Thomeeque]

OK, I have made it real..



..then I have made some tweaks (e.g. buffer for BBD output)..



..and now I have question, which I will start new thread about

 T.

[Thomeeque]

..and now I have question, which I will start new thread about

 BBD (TDA1022) - Tetrode Gate - correct voltage?

[oldschoolanalog]

How does it sound?

[e45tg4t3]

Nice Build^^... would you mind to share the layout?

Best Regards

Ben

[Thomeeque]

How does it sound?

 Thanks for your interest! So far I'd say it sounds promising I'm still in the middle of tweaking (actual biggest problem is relatively big /around 6dB/ attenuation of TDA1022 compared to original SAD1024 or actual RD1506, maybe I'll have to redesign the BBD buffer to make some boost), but I'm already getting very nice flanging with typical EM sweep and similar range of useful sounds, so I'm optimistic.

Nice Build^^... would you mind to share the layout?

Best Regards

Ben

 Thanks! I'm planning to share the layout, but I'd wait till the tweaking is done (let me know, if you don't agree )

[Thomeeque]

 Hi!

 Here's the latest status:

 

 It's mostly me vs. TDA1022's attenuation fight these days, first I have replaced load resistor by current source, which lowers attenuation approximately from 6 to 2.5dB (~ increases gain from 0.5 to 0.75) and now I'm on hunt for low-noise non-inverting boosting buffer, which will move it to SAD1024's gain level (SAD1024 gain is around 1.1), maybe I'll start another thread again

 Btw. to by able to nail original EM character, I have created semi-interactive (interactive by altering code ) AC transitions emulation.

 Here's what I've got for 9V EM (it drops signal down by cca 10dB!, buyers must have really loved the effect produced when they accepted this ):



 And here's for comparison actual Deluxe EM (from factory schematic; buffer trimmer is "set" for maximal signal) (they cared to keep approx. unity gain here ):



Feedback represents transition between points 4 and 2 via feedback path..

Unwarranted, it can be all wrong

[Thomeeque]

 Hi!

 So it's sort-of* finished

 

 More pics here.

 Since buffers were inevitable, I had decided to take advantage of this fact and added few new features to the circuitry (polarity switches, vibrato mode, chorus mode..).

 Final circuitry looks like this:

 

 

 I would share PCB loyout as well, but unfortunately I have made lot of changes (mainly on main board), so it's obsolete now and needs huge update.

 Honestly, I'm not even sure it's so good idea to build this anyway (it's a bit noisy, but maybe it's normal for all BBD based circuits?), maybe wait for some sound samples (which I plan to create) before deciding..

 I'm bit short of time now, but if you have questions just ask, I'll answer soon or later.. T.

* I may still try some tuning (when I collect some energy again ), e.g. increasing of gain on input buffer to get better SNR and maaaaaybe clock buffering (but original EM does not do that as well, so maybe not)

EDIT: Btw. thanks to markusw for his TZF Electric Mistress (sample and layout) thread, I've loaded some inspiration from there

[oldschoolanalog]

Nice work! Thanks for doing all this and sharing your efforts.
I patiently await the sound samples.
One Q. What are the highest & lowest clock f's this is capable of (measured at the BBD please) when in "filter matrix" mode?
Thanks again ,
Dave