A/DA flanger, some practical questions.

[Dirk_Hendrik]

Since I recently gained interest in doing a version of the A/DA flanger I started wondering and hope someone here is able to answer the following:

A MN1010 is used. I have one, harvested from an old VCR years ago, so no immediate problems but still wondering:
- The schematic states clock frequencies ranging from 17.4 to 650 Khz. Checking the MN3010 datasheet (BBDementia, thanks Mark) states a max clock of 100 khz. This implies serious overclocking....?
- The MN3010 contains two 512 stage delay lines connected in parallel. The only reason I can come up with for doing this is an improved SN ratio. Any other explanations?
- In the last case, should't 2 MN3002 BBD's give the same results?

Especailly the first implies a problem since I was hoping to double the clock and use a (or 2 in parallel)  cheap BL3207 (with the neccessary changes) and have some more room for screwing up.

[Dirk_Hendrik]

:shock:

no-one?

[Bernardduur]

He Nederlander,

I read in the manual that comes with the flanger that the clock frequency had to be calibrated and the only two speeds I find there are 34.8 kHz (Bias trimpot T1) and 1.3 kHz (Bias trimpots T2 and T4 and watch for interference between the two).

The sweep then should be between 1.3 and 34.8 (if not adjust bias tripot T3)....

Dunno if it is the same chip, but this is the only clock frequency they mention.

[Dirk_Hendrik]

mogge man,

Read again :wink:  That's not 1.3 Khz but 1300 kHz. The 650 kHz I was referencing to is after the MN driver which gives  division of the master clock by 2.

[Dirk_Hendrik]

Re-check.
Apparently there's 2 versions of this calibration procedure. One say 1300Khz and the other 1.3 Khz. Looks like an ADA typo.

[Bernardduur]

LOL

I saw it. So it suppose to be 1.3 MHz.....

[Dirk_Hendrik]

Guess not. Guess it's the other way around and they should've ommitted a "k".
Nevertheless,
Since the maxclock is some 40 khz and the max allowed clock 100kHz it is interesting to redesign the clock to produce a double frequency and use cheap BL/MN3207's.

[Mark Hammer]

One of the other advantages gained by using two 512-stage sections in parallel is that the effective delay time achieved is cut in half, in comparison to 1024 stages in series.  Since the most dramatic flange sounds (and part of what the A/DA is known for too) come when the unit is capable of approaching zero delay, the dual-parallel arrangement makes it easier to reach even shorter delays.  Mike Irwin explained to me that one of the common obstacles to overcome in BBDs is the input capacitance of the clock pins on the BBD.  All BBDs have this, and it tends to "round off" the edges of clock pulses at higher clock rates.  It can be overcome by buffering the clock lines (one of the reasons you will often see a couple of parallel inverters used between the clock source and the clock input pins on a BBD), but lower you can *afford* to drive the clock to achieve the desired delay times, the easier the job will be.  Using two parallel 512-stage units rather than one 1024-stage unit makes the job easier.

[DiyFreaque]

Another cool trick with using two BBD's is configuring them for the parallel multiplex mode - using two individual BBD's to effectively double your resolution with the same clock frequency.  It's explained very well in the SAD1024 datasheet.  

It works basically like this:  most BBD signals have two outputs (one exception being the MN3011 tapped outputs).  The dual clock waveform puts out one sample on each output.  On a single BBD, the first of the dual clock pulses puts out a sample on one output, and the second clock pulse puts out the *same* sample on the other output when it goes high.  Those two outputs are then mixed together to get your delayed signal.  

With parallel mutliplexing, the idea is to use two different BBD sections (or two BBD's) and use only one output of each BBD.  On one BBD you wire the clock inputs so that one pulse of the two phase clock chucks the first sample out on the output you're using.  On the second BBD, the clock inputs are 'reversed' so you chuck the first sample of that BBD out with the 'second' clock pulse of the two phase clock.  So, for each clock cycle of the two phase clock you're putting out two individual 'fresh' samples rather than one sample and a copy of that sample, effectively doubling your sampling rate with the same clock frequency.  IIRC, Mike has that arrangment on his A/DA clone.  

Cheers,
Scott

[Dave_B]

Can someone tell me what the MN1010 chip does?  I've got a couple of limping VCR's that may have some harvesting in their future.

...and yes, I searched the archive first.   

[Mark Hammer]

A quick Google search indicates the MN1010 is a GPS chip, and not related to either video recorders OR rock and roll.  Are you sure the '1' is not a '3' with some paint scratched off?

If you look at some of the datasheets for some of the smaller-capacity BBDs (e.g., 128 stage) and for the Reticon chips, one of the applications is for syncing of video signals.  So finding a BBD in a VCR somewhere is not completely unreasonable. 

On the other hand, do not assume that the prefix "MN" is necessarily associated ONLY with BBDs.  Companies are also free to have their own house numbers for components, and stick an MN prefix on them if they so choose.

[cd]

Can someone tell me what the MN1010 chip does?  I've got a couple of limping VCR's that may have some harvesting in their future.

...and yes, I searched the archive first.   

I think it was just a typo.

[Dirk_Hendrik]

Alarm Alarm.
Ignore the typo! Another user pointed that out to me already. I was referring to a MN3010 indeed. Furthermore, the populated PCB was tested successfully 2 weeks ago.

In the VCR this one came from (I guess it was an old Blaupunkt CHS and definitely a toploader) the MN was in the audio circuitry for sure. The harvesting was some 10 years ago already. At that time, with the absence of Matsuhitamania/bbdementia, it was close to impossible to obtain datasheets for MN's. The only reason for keeping the PCB containing this chip was because it said MN3xxx ans rang a bell as "could be BBD, could be expensive chip". I only dug that PCB up from te crap crate when I found that full design file for an ADA flanger.

The MN prefix is pretty common for Matsuhita/Panasonic but indeed not limited to BBD's. The same flanger "proto" which is on my workbench still but now fully populated shows one MN chip:


In this case an MN4007. Definitely not BBD. Standard CMOS logic gate.