[sdiy] reducing crosstalk

Thu Feb 21 14:40:37 CET 2008

Dear list,

thanks to everyone for your thoughts on this project. The really 
exciting part of this whole DIY electronics world is to get feedback 
like all of yours, it really gives me a lot of input to improve my 
designs in the future!

I realized that the biggest problem was the current to the 
phototransistors. In my original, single-supply box with no buffers, I 
used a single, standard 5 VDC regulator construction with filtering 
capacitors sent to all 8 phototransistors, without problem. When I made 
the synth module, somehow I thought that a resistance voltage divider 
(yes, drawn incorrectly in the schematic but soldered correctly on the 
board) would be enough. But impedance matching is still a new topic for 
me, so I didn't realize that there would be such a huge voltage drop on 
all the channels, which caused the crosstalk.

The immediate solution to my problem with simply to replace the voltage 
divider with a proper 5V voltage regulator and some filtering capacitors 
on the flow to all the phototransistors. If that had failed, I would 
have tried filtering the +/- supply to the ICs with 0.1 uF bypass caps, 
and after that to try to properly regulate the flow to each 
phototransistor. Luckily, the first improvement solved the problem.

Big thanks and best wishes,
Derek

http://www.umatic.nl/tonewheels.html

Derek Holzer wrote:
> Hi Neil,
> 
> thanks to you and also to the other members of this list for the 
> replies. Perhaps I drew it wrong, but on the board it really is a 
> voltage divider! ;-)
> 
> The two stages are because each op-amp is inverting. I tried a single 
> non-inverting, but the input impedance was far too high to get anything 
> out. I don't want to filter DC because I use it in two ways, one being 
> direct optical synthesis and the other being control voltage for a 
> modular synth. Also for this reason, the output voltage should be 
> positive. Later on I realized that the voltage to the phototransistor 
> could have been negative, and then a single inverting stage would be 
> fine. But I started this project on a single supply, and then moved to 
> dual supply, so some old thinking came along for the ride.
> 
> The first change I will make (and actually what I was doing when it was 
> a single-supply box that didn't have any crosstalk) will be to use a 5V 
> voltage regulator to the BPW40s. And, as several have mentioned, I'll 
> decouple the supply to the ICs. Should happen this weekend, so I'll be 
> happy to report back the results and post a new diagram.
> 
> Also, I realized I didn't send the link to the project itself, could be 
> interesting for folks here interested in exotic synthesis techniques, 
> espc the HISTORY section of the website:
> 
> http://www.umatic.nl/tonewheels.html
> 
> thx + best!
> d.
> 
> Neil Johnson wrote:
>> Hi,
>>
>> Looking at that circuit, it is not a voltage divider.  All it is a 10k 
>> load across the +12V supply, and a 10k load resistor for the 
>> phototransistor.  You might want to swap the connections to +12V and 
>> the BPW40 to fix that if you really want to drive the BPW40 from 6V.  
>> As Alan suggests, a low-impedance source, like a voltage regulator or 
>> another op-amp would sort out that.
>>
>> Why do have two stages of unity gain?  Surely for this application it 
>> doesn't matter?  A single stage would be sufficient.
>>
>> I'd also put a DC blocking capacitor on the output, plus a series 
>> resistor of, say, 100R.
>>
>> For each module I'd put some filtering on the incoming supply, to help 
>> isolate each module.
>>
>> Put 100n decoupling caps close to the supply pins on the TL074.
>>
>> There is nothing of high gain here so shielding cans are probably not 
>> needed.
>>
>> Neil
>>
>> On 14 Feb 2008, at 17:16, Needham, Alan wrote:
>>
>>>
>>>
>>>> -----Original Message-----
>>>> From: synth-diy-bounces at dropmix.xs4all.nl
>>> [mailto:synth-diy->bounces at dropmix.xs4all.nl] On Behalf Of Derek Holzer
>>>> Sent: 25 January 2008 20:01
>>>> To: synth-diy at dropmix.xs4all.nl
>>>> Subject: [sdiy] reducing crosstalk
>>> <snip>
>>>> There is crosstalk between all the channels, not just the channels
>>>> Which share an IC.
>>> <snip>........................................
>>>
>>> Derek, could the above be the key here?
>>> What is shared between all channels,
>>> You say there is some 0.1u caps across the power rails, try adding a
>>> couple of 10u or 100u caps as smoothing (I never did care for specific
>>> values - whatever you have should prove a point). If the crosstalk
>>> reduces then we may be onto something...
>>> ...BUT, more likely...
>>> Also shared between all channels (as I read your diagram) is the two
>>> resistor potential divider for the 6v feed to all optos. This seems to
>>> be two 10k resistors, the input impedance to each inverting amp is also
>>> 10k (all values read from photos - so am I right?).
>>> The impedance of this 6v rail seems to be too high for the 8 amps it is
>>> expected to drive. Lash a 100u or 470u cap from 6v to 0v and see if that
>>> helps, if it does, you could also try swapping the first amp stage's
>>> input resistor and feedback resistor for 100k on all 8 channels, 16
>>> resistors, or does this wreck the opto's operation???
>>> Ideally, the 6v rail would be a regulator (78L05?) with bypass caps.
>>>
>>>     Alan
>>>
>>>
>>> _____________________________________________________________________
>>> The information contained in or attached to this email is intended 
>>> only for the use of the individual or entity to which it is 
>>> addressed. If you are not the intended recipient, or a person 
>>> responsible for delivering it to the intended recipient, you are not 
>>> authorised to and must not disclose, copy, distribute, or retain this 
>>> message or any part of it. It may contain information which is 
>>> confidential and/or covered by legal professional or other privilege 
>>> (or other rules or laws with similar effect in jurisdictions outside 
>>> England and Wales).
>>> The views expressed in this email are not necessarily the views of 
>>> Centrica plc, and the company, its directors, officers or employees 
>>> make no representation or accept any liability for its accuracy or 
>>> completeness unless expressly stated to the contrary.
>>>
>>> Centrica plc
>>>
>>> Registered office: Millstream, Maidenhead Road, Windsor, Berkshire 
>>> SL4 5GD
>>>
>>> Registered in England and Wales No 3033654
>>>
>>> _______________________________________________
>>> Synth-diy mailing list
>>> Synth-diy at dropmix.xs4all.nl
>>> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>>
>> -- 
>> http://www.njohnson.co.uk
>>
>>
>>
>>
> 

-- 
derek holzer ::: http://www.umatic.nl ::: http://blog.myspace.com/macumbista
---Oblique Strategy # 146:
"Spectrum analysis"