Hi ErikdeBest,
No problem, it wasn't your fault.
Yes, it's the same.
Due to specific diode properties (Schottky diode has much higher leakage current than the HV silicon diode), almost all reverse voltage will be across the fast-slow recovery silicon HV diode. So basically no equalization should be required.
One important note though, when using Schottky (rectifier) diodes, make sure they don't run hot (use heatsink when required). Temperature increase will increase leakage current, this means that the diodes no longer fully block the reverse voltage, and still pass a small amount of current.
The BAT42 would survive (470K series resistor limits current to a safe value), but a high voltage version (hybrid diode) is required for HV applications.
The voltage between both gate and source is limited by either internal or external zener diodes. If the HV MOSFET doesn't have built-in gate protection diodes, these have to be added externally, as described in post #2347. Vgs shouldn't exceed +/- 12 ... 20V depending on protection diode value.
The BAT42 is added to switch-off the MOSFET fast, and switch-ON slow. When the input voltage is positive (positive power supply) the gate will be pulled high through 100K in series with the BAT42 (BAT42 conducts), so basically by 100K. When the input voltage drops, the gate will be pulled low with increasing voltage (as the sine wave amplitude drops), now the BAT42 blocks the current, and the gate has to be pulled low through 100K + 1 M Ohm = 1.1 M Ohm, providing slow turn-ON. This prevents peak charge-transfer current (there is plenty of time to transfer the charge between both caps).
2SJ306 (P), 2SJ449 (P), 2SK1119 (N), FQP1P50 (P), FQP3P50 (P), MTP2P50EG (P), FQPF3P50 (P), IRFI9634GPBF (P), FQP4P40 (P).
No problem, it wasn't your fault.
Yes, it's the same.
Due to specific diode properties (Schottky diode has much higher leakage current than the HV silicon diode), almost all reverse voltage will be across the fast-slow recovery silicon HV diode. So basically no equalization should be required.
One important note though, when using Schottky (rectifier) diodes, make sure they don't run hot (use heatsink when required). Temperature increase will increase leakage current, this means that the diodes no longer fully block the reverse voltage, and still pass a small amount of current.
The BAT42 would survive (470K series resistor limits current to a safe value), but a high voltage version (hybrid diode) is required for HV applications.
The voltage between both gate and source is limited by either internal or external zener diodes. If the HV MOSFET doesn't have built-in gate protection diodes, these have to be added externally, as described in post #2347. Vgs shouldn't exceed +/- 12 ... 20V depending on protection diode value.
The BAT42 is added to switch-off the MOSFET fast, and switch-ON slow. When the input voltage is positive (positive power supply) the gate will be pulled high through 100K in series with the BAT42 (BAT42 conducts), so basically by 100K. When the input voltage drops, the gate will be pulled low with increasing voltage (as the sine wave amplitude drops), now the BAT42 blocks the current, and the gate has to be pulled low through 100K + 1 M Ohm = 1.1 M Ohm, providing slow turn-ON. This prevents peak charge-transfer current (there is plenty of time to transfer the charge between both caps).
2SJ306 (P), 2SJ449 (P), 2SK1119 (N), FQP1P50 (P), FQP3P50 (P), MTP2P50EG (P), FQPF3P50 (P), IRFI9634GPBF (P), FQP4P40 (P).
Hi PyroVeso,
Should work fine, it also has intergrated protection diodes for the gate. The capacitance is no problem as a 11DQ10 Schottky diode is placed in series with the drain.
The IRL540N doesn't seem to have integrated protection diodes for the gate, so these would need to be added externally.
The MOSFETs aren't very critical, so many types with similar specs can be used.
Should work fine, it also has intergrated protection diodes for the gate. The capacitance is no problem as a 11DQ10 Schottky diode is placed in series with the drain.
The IRL540N doesn't seem to have integrated protection diodes for the gate, so these would need to be added externally.
The MOSFETs aren't very critical, so many types with similar specs can be used.
Hi Rick Miller,
Yes I tested bipolar transistors (darlingtons) for both the capacitance multiplier, and charge-transfer circuit. They work too, but are usually less efficient (larger voltage drop).
Both resistor and capacitor values for driving the bases of the darlingtons need to be adapted,
I will give them a try.
Yes I tested bipolar transistors (darlingtons) for both the capacitance multiplier, and charge-transfer circuit. They work too, but are usually less efficient (larger voltage drop).
Both resistor and capacitor values for driving the bases of the darlingtons need to be adapted,
I will give them a try.
Hi JC951t,
I really like how Teddyregs "sound". I have not tested other cap combination, save putting film caps parallel tants on C6.
I trusted the inventor's opinion about why they were preferable and stuck with them:
Please, take a look at:
http://www.pinkfishmedia.net/forum/showthread.php?t=39990
Cheers,
M
I really like how Teddyregs "sound". I have not tested other cap combination, save putting film caps parallel tants on C6.
I trusted the inventor's opinion about why they were preferable and stuck with them:
Please, take a look at:
http://www.pinkfishmedia.net/forum/showthread.php?t=39990
Cheers,
M
I have found that adding a simple RC after the main filter cap really reduces the ripple and high freq noise that pass thru the supply from the mains. The R can be sized so that you don't have too much voltage drop. I use between 2-10 ohms for the R and 2200 - 4700uf for the C. You can see with your 'scope how much it reduces the ripple and with a roll off of 6db/oct think what it does to the very high freq noise.
Hi Rick Miller,
Yes this should work fine, I put resistors of 0.5 ... 1 Ohm in series with each rectifier diode. It also reduces ripple voltage and RF / HF noise. Disadvantage is extra voltage drop. R4 serves a similar purpose.
Some tips about the charge-transfer power supply:
Add 330 Ohm between gate and R5 / C3, this will prevent possible spurious oscillations with some MOSFETs. Keep the connection between both gate and 330 Ohm resistor as short as possible.
For HV charge transfer power supplies:
- Remove R2 / D5, R1 (100K) goes directly to the gate of T1.
- Change R3 to 1 M.
Some HV MOSFETs don't seem to work properly in the capacitance multiplier (too high ripple voltage), use HV darlington instead like BU806. R5 then needs to be lowered to approx. 10K (I use a darlington in the 200V anode power supply). Also put a small series resistor (approx. 220 Ohm) between base and R5 / C3.
Just found a very low capacitance Schottky diode for D4 (datasheet indicates Cj = 10pF, 11DQ10 is specified at approx. 55pF). Type number SK26AR2, 60V / 2A, available from RS (P/N. 652-6053).
When constructing low voltage supplies, 11DQ10 (rectifier diodes) could be replaced with 11DQ6 (lower voltage drop).
Yes this should work fine, I put resistors of 0.5 ... 1 Ohm in series with each rectifier diode. It also reduces ripple voltage and RF / HF noise. Disadvantage is extra voltage drop. R4 serves a similar purpose.
Some tips about the charge-transfer power supply:
Add 330 Ohm between gate and R5 / C3, this will prevent possible spurious oscillations with some MOSFETs. Keep the connection between both gate and 330 Ohm resistor as short as possible.
For HV charge transfer power supplies:
- Remove R2 / D5, R1 (100K) goes directly to the gate of T1.
- Change R3 to 1 M.
Some HV MOSFETs don't seem to work properly in the capacitance multiplier (too high ripple voltage), use HV darlington instead like BU806. R5 then needs to be lowered to approx. 10K (I use a darlington in the 200V anode power supply). Also put a small series resistor (approx. 220 Ohm) between base and R5 / C3.
Just found a very low capacitance Schottky diode for D4 (datasheet indicates Cj = 10pF, 11DQ10 is specified at approx. 55pF). Type number SK26AR2, 60V / 2A, available from RS (P/N. 652-6053).
When constructing low voltage supplies, 11DQ10 (rectifier diodes) could be replaced with 11DQ6 (lower voltage drop).
The voltage drop can become hearable. Swapped the regular 15A rectifiers by diy ones with HFA08TB60 hexfred diodes in the fet-section of hybrid amp. To my taste it sounded way to laidback. Replaced the hexfreds with big TO220 BYV79E diodes, and that sounded a lot better. So for poweramps or schematics with larger currents a rect. diode can be chosen with a low F-voltage drop. In HT PS it is the low voltage drop of no concearn.
An externally hosted image should be here but it was not working when we last tested it.
Hi maxlorenz,
The Teddyreg has limited bandwidth for corrections (LC filter), so the output cap C6 needs to provide a low impedance for higher frequencies. This means it probably has a big effect on how the circuit performs. If the tantalium cap + X7R ceramic cap provide flat frequency response, it should work fine.
Today I built the Teddyreg (again), testing 45DH11 and the 2SK117 JFET combination. It had significant voltage drop (load current 266mA), and it didn't seem to suppress the ripple voltage. Performance was much better when using a darlington (2SC3423 + 2SC2240). The low voltage MOSFET version provided best ripple voltage suppression (down to approx. 100uV rms), while providing lowest voltage drop at 266mA load current (only 1.7V). MOSFETs have the advantage of very low Rdson (low losses), and almost infinite current gain.
I have to add that I use the capacitance multiplier in a different way than with the Teddyreg (it's placed in front of the ultra-low noise voltage regulators). This is done to use local voltage regulators that provide a very stable voltage, close to the circuits where they are needed. The capacitance multiplier ensures that the voltage regulators run on clean DC voltages (with very low ripple voltage). This appears to result in much better regulator performance. When a voltage regulator runs on a polluted power supply with high ripple voltage and noise, it has to compensate for both, the polluted input voltage, and load current fluctuations. Now the voltage regulator mainly needs to compensate for load current fluctuations only.
I try to keep HF loops as short as possible (local loops). This means that there shouldn't be a ceramic cap in the capacitance multiplier that feeds the local voltage regulators and connected circuits. C4 won't provide a very low impedance for HF, opening the long loop, but with the 330R resistor preventing MOSFET spurious oscillations, it should be possible to remove C4 as well. This will force all LF ... HF frequencies to take local loops, preventing unwanted inter-modulation on the power supply distribution wires / traces.
Voltage regulation is done locally (ultra-low noise regulators), and HF decoupling is also done locally (very close to the circuits that produce HF noise), I also use tantalium caps (SMD) with ceramic X7R or NPO caps for this purpose.
The Teddyreg has limited bandwidth for corrections (LC filter), so the output cap C6 needs to provide a low impedance for higher frequencies. This means it probably has a big effect on how the circuit performs. If the tantalium cap + X7R ceramic cap provide flat frequency response, it should work fine.
Today I built the Teddyreg (again), testing 45DH11 and the 2SK117 JFET combination. It had significant voltage drop (load current 266mA), and it didn't seem to suppress the ripple voltage. Performance was much better when using a darlington (2SC3423 + 2SC2240). The low voltage MOSFET version provided best ripple voltage suppression (down to approx. 100uV rms), while providing lowest voltage drop at 266mA load current (only 1.7V). MOSFETs have the advantage of very low Rdson (low losses), and almost infinite current gain.
I have to add that I use the capacitance multiplier in a different way than with the Teddyreg (it's placed in front of the ultra-low noise voltage regulators). This is done to use local voltage regulators that provide a very stable voltage, close to the circuits where they are needed. The capacitance multiplier ensures that the voltage regulators run on clean DC voltages (with very low ripple voltage). This appears to result in much better regulator performance. When a voltage regulator runs on a polluted power supply with high ripple voltage and noise, it has to compensate for both, the polluted input voltage, and load current fluctuations. Now the voltage regulator mainly needs to compensate for load current fluctuations only.
I try to keep HF loops as short as possible (local loops). This means that there shouldn't be a ceramic cap in the capacitance multiplier that feeds the local voltage regulators and connected circuits. C4 won't provide a very low impedance for HF, opening the long loop, but with the 330R resistor preventing MOSFET spurious oscillations, it should be possible to remove C4 as well. This will force all LF ... HF frequencies to take local loops, preventing unwanted inter-modulation on the power supply distribution wires / traces.
Voltage regulation is done locally (ultra-low noise regulators), and HF decoupling is also done locally (very close to the circuits that produce HF noise), I also use tantalium caps (SMD) with ceramic X7R or NPO caps for this purpose.
Thanks -EC- for your explanations and advices.
I use mainly Darlington (44H11 + BC550) based gyrator. I'll keep on using TeddyReg's (not that I'm stubborn ) because it's performance is satisfactory (until now) for me and because I am familiar with it, apart the fact that my limited knowledge makes me predict more troubles than benefits if I use other approach...
I could test Mosfet gyrator...
I also use local HF decoupling with Tant + X7R (SMD) next to the chips.
How many low noise regulators will you use on your boards?
Cheers,
M
PS: I am about to finish the DI4M with scrambler...wish me luck.
I use mainly Darlington (44H11 + BC550) based gyrator. I'll keep on using TeddyReg's (not that I'm stubborn
) because it's performance is satisfactory (until now) for me and because I am familiar with it, apart the fact that my limited knowledge makes me predict more troubles than benefits if I use other approach...I could test Mosfet gyrator...
I also use local HF decoupling with Tant + X7R (SMD) next to the chips.
How many low noise regulators will you use on your boards?
Cheers,
M
PS: I am about to finish the DI4M with scrambler...wish me luck.
Hi maxlorenz,
I currently have one power supply PCB containing 3 charge-transfer circuits, and 3 MOSFET capacitance multipliers. The power supply outputs -18, -7.5 and +7.5V DC (capacitance multiplier outputs).
These power supplies are then routed to the DI4T mainboard. I use 4 x Micrel (programmable) ultra low noise regulators (MIC5205-5.0YM5) to drive the logic circuits. The ultra low noise regulator outputs are filtered (passive LC filters) to further attenuate noise. The masterclock has a special LC filter (40 Henry hybrid inductor) that's placed in a mu-metal housing.
The masterclock (on-board) power supply decoupling includes a supercap, tantalium caps, X7R ceramic caps, and an additional HF LC filter. I can no longer measure any jitter with my measuring instruments, I do hear slight sound quality degradation when I put a HF probe on the masterclock clock buffer output for measuring.
I still use selected 78xx and 79xx regulators for the TDA1541A power supplies, and they seem to work very well when running on a clean, almost ripple-free DC input voltage. However, I might use capacitance multipliers with integrated voltage regulator, these could provide +5V, -5V and -15V ultra low noise power supplies for the TDA1541A chips, then all voltage regulators on the DA1541A modules could be left-out.
The tube power supply now contains 2 charge-transfer circuits, and 2 capacitance multipliers (-40V, and +200V). The filaments still run on a conventional 12.6V DC power supply (7812 + silicon diode to get 12.6V output). Both -40V and +200V are filtered using 10H chokes. The differential input stage has an extra filter.
Yesterday I tested a twin charge-transfer power supply to drive 2 groups of 2 filaments (2 x 12.6V / 300mA) directly from stabilized capacitance multipliers. It works fine, and allows for floating filament power supplies (2 separate secondary windings with rectifier bridges). This is useful to keep the voltage potential between both cathode and filaments as low as possible.
I wish you luck with the DI4M, and I am curious about the results.
I currently have one power supply PCB containing 3 charge-transfer circuits, and 3 MOSFET capacitance multipliers. The power supply outputs -18, -7.5 and +7.5V DC (capacitance multiplier outputs).
These power supplies are then routed to the DI4T mainboard. I use 4 x Micrel (programmable) ultra low noise regulators (MIC5205-5.0YM5) to drive the logic circuits. The ultra low noise regulator outputs are filtered (passive LC filters) to further attenuate noise. The masterclock has a special LC filter (40 Henry hybrid inductor) that's placed in a mu-metal housing.
The masterclock (on-board) power supply decoupling includes a supercap, tantalium caps, X7R ceramic caps, and an additional HF LC filter. I can no longer measure any jitter with my measuring instruments, I do hear slight sound quality degradation when I put a HF probe on the masterclock clock buffer output for measuring.
I still use selected 78xx and 79xx regulators for the TDA1541A power supplies, and they seem to work very well when running on a clean, almost ripple-free DC input voltage. However, I might use capacitance multipliers with integrated voltage regulator, these could provide +5V, -5V and -15V ultra low noise power supplies for the TDA1541A chips, then all voltage regulators on the DA1541A modules could be left-out.
The tube power supply now contains 2 charge-transfer circuits, and 2 capacitance multipliers (-40V, and +200V). The filaments still run on a conventional 12.6V DC power supply (7812 + silicon diode to get 12.6V output). Both -40V and +200V are filtered using 10H chokes. The differential input stage has an extra filter.
Yesterday I tested a twin charge-transfer power supply to drive 2 groups of 2 filaments (2 x 12.6V / 300mA) directly from stabilized capacitance multipliers. It works fine, and allows for floating filament power supplies (2 separate secondary windings with rectifier bridges). This is useful to keep the voltage potential between both cathode and filaments as low as possible.
I wish you luck with the DI4M, and I am curious about the results.
Hi Rick Miller,
I not published any schematics yet.
I just completed PCB design of two new discrete voltage regulators I am planning to use in my DACs.
The regulators measure approx. 10 x 21 x 6mm (almost the same size as conventional 78xx and 79xx regulators)
There are versions for both positive and negative supplies, they are pin-compatible with both 78xx and 79xx regulators. I will start with +5V, -5V, and -15V.
The discrete SMD miniature regulators have an unusual design. The low noise bandgap reference diode is running on the stabilized output voltage.
The reference output is filtered (noise) and drives an OP-amp with variable gain. The output voltage is divided by a resistive divider, this enables changing output voltage to desired values. The series element is a darlington.
The OP-amp runs on a miniature MOSFET gyrator, reducing input power supply voltage ripple below approx. 1mV rms. This ensures very stable Op-amp operation.
I not published any schematics yet.
I just completed PCB design of two new discrete voltage regulators I am planning to use in my DACs.
The regulators measure approx. 10 x 21 x 6mm (almost the same size as conventional 78xx and 79xx regulators)
There are versions for both positive and negative supplies, they are pin-compatible with both 78xx and 79xx regulators. I will start with +5V, -5V, and -15V.
The discrete SMD miniature regulators have an unusual design. The low noise bandgap reference diode is running on the stabilized output voltage.
The reference output is filtered (noise) and drives an OP-amp with variable gain. The output voltage is divided by a resistive divider, this enables changing output voltage to desired values. The series element is a darlington.
The OP-amp runs on a miniature MOSFET gyrator, reducing input power supply voltage ripple below approx. 1mV rms. This ensures very stable Op-amp operation.
DI4JM: scrambler and everything.
Hey John, do you want to see my modular configuration?
You could learn a few things...
http://picasaweb.google.com/lh/photo/PZfaGSeS0YuWlKVUDImcsg?feat=directlink
(3 picks; use magnifying lens option)
(Yeah...I overestimated the size of the box )
Cheers,
M
Hey John, do you want to see my modular configuration?
You could learn a few things...
http://picasaweb.google.com/lh/photo/PZfaGSeS0YuWlKVUDImcsg?feat=directlink
(3 picks; use magnifying lens option)
(Yeah...I overestimated the size of the box
)Cheers,
M
Hi John, I'm sure your Dac sounds great.
I'm in need of a great Dac and this monolith thread has my attention
It seems a couple of the experience DAC builders have built your basic design already.. but being somewhat of a novice in electronics , I hope a finalized design is re-presented so that myself (or dumb ***** like me) will have a chance at attempting.
As for now I can only dream for a kit and a builders thread
keep up the good work! and Would love to hear more on how it sounds!!
- pls help me ditch my benchmark dac1!!
I'm in need of a great Dac and this monolith thread has my attention
It seems a couple of the experience DAC builders have built your basic design already.. but being somewhat of a novice in electronics , I hope a finalized design is re-presented so that myself (or dumb ***** like me) will have a chance at attempting.
As for now I can only dream for a kit and a builders thread
keep up the good work! and Would love to hear more on how it sounds!!
- pls help me ditch my benchmark dac1!!
Charge-Transfer Power Supply
Hi ecdesign,
I need your advise.
I built your, the excellent idea, charge-transfer power supply as of schematic on post #2347, intended to test with my shigaclone cd transport.
I built only charge-transfer without capacitance multiplier. I used 2SJ176 (same pin outline of 2SJ380) in substitute of 2SJ380 and 1N4937 in replace of all diodes. That's only I can find in local market here. Both C are 2000uF. After second C, I used TeddyReg with 1000uF output C, then supply to Shigaclone. The transformer secondary can supply 3A contineuos without heat.
It's not work. The display shown -00- but the spindle is not work. I measured 13.8V on first C and 12.8V before TeddyReg and 8.5V after TeddyReg. (Shigaclone can stand on 8-10V without problem)
Next, I changed TeddyReg to LM7808 and got 8.0V after regulator. Same result.
Next, I added 10000uF to first C in charge-transfer set. Then, I added 10000uF to second C. I remove D4 in order to reduce V drop. All these trying are not work.
I bypassed the charge-transfer set, then Shigaclone back to work normally.
Could you please advise me what I should do next? Thank you very much.
Regards,
Art.
Hi ecdesign,
I need your advise.
I built your, the excellent idea, charge-transfer power supply as of schematic on post #2347, intended to test with my shigaclone cd transport.
I built only charge-transfer without capacitance multiplier. I used 2SJ176 (same pin outline of 2SJ380) in substitute of 2SJ380 and 1N4937 in replace of all diodes. That's only I can find in local market here. Both C are 2000uF. After second C, I used TeddyReg with 1000uF output C, then supply to Shigaclone. The transformer secondary can supply 3A contineuos without heat.
It's not work. The display shown -00- but the spindle is not work. I measured 13.8V on first C and 12.8V before TeddyReg and 8.5V after TeddyReg. (Shigaclone can stand on 8-10V without problem)
Next, I changed TeddyReg to LM7808 and got 8.0V after regulator. Same result.
Next, I added 10000uF to first C in charge-transfer set. Then, I added 10000uF to second C. I remove D4 in order to reduce V drop. All these trying are not work.
I bypassed the charge-transfer set, then Shigaclone back to work normally.
Could you please advise me what I should do next? Thank you very much.
Regards,
Art.
Hi Art,
Sorry about your problem.
Note that the classical TeddyReg could be tricky or not work at all with circuits in which the timing of the power up, as maybe is the case of your transport, is critical. This is caused because the Vout takes several seconds to build-up. Teddy Pardo (the inventor) posted an "accelerated" version @pinkfishmedia...
As for the charge-transfer circuit it is better that -EC- answers
The DI4M is finally in working condition. This was the most challenging project that I've done so far, but I think that now I could make another in an easier way. I had some troubles like an USB receiver blown-up (my mistake; never desolder while USB cable is connected
) and a little short of one DAC output pin ...
We live and learn.
I'll try to post some more picks today from the scope.
No critical or comparative listening yet...
Thanks -EC-
M.
Sorry about your problem.
Note that the classical TeddyReg could be tricky or not work at all with circuits in which the timing of the power up, as maybe is the case of your transport, is critical. This is caused because the Vout takes several seconds to build-up. Teddy Pardo (the inventor) posted an "accelerated" version @pinkfishmedia...
As for the charge-transfer circuit it is better that -EC- answers
The DI4M is finally in working condition. This was the most challenging project that I've done so far, but I think that now I could make another in an easier way. I had some troubles like an USB receiver blown-up (my mistake; never desolder while USB cable is connected
) and a little short of one DAC output pin ...We live and learn.
I'll try to post some more picks today from the scope.
No critical or comparative listening yet...
Thanks -EC-
M.
My Ultimate NOS DAC with TDA1543.
Some picks from the scope: digital chain and balanced signal...
BCK and DATA:
http://picasaweb.google.com/lh/photo/QWGAgdw27xCo2IjZQmQ0qA?feat=directlink
The picks are not so good because I had both hands on the probes
My "assistant" is an excellent (and I mean excellent) cook but a poor photographer...
WS and DATA from non-inverting DAC:
http://picasaweb.google.com/lh/photo/Byy5T1Q8p9oJH9Unycz48w?feat=directlink
WS and DATA from inverting DAC:
http://picasaweb.google.com/lh/photo/PiMieIoOk2bq3gCC68xkSg?feat=directlink
Input signal from parallel DACs to differential opamp, left channel:
http://picasaweb.google.com/lh/photo/v-_UyMzDMTCyduKwAniZRQ?feat=directlink
The same, right channel:
http://picasaweb.google.com/lh/photo/K-bqf5sv76zRZhWwwY4V9Q?feat=directlink
I hope you like it
Cheers,
M
Some picks from the scope: digital chain and balanced signal...
BCK and DATA:
http://picasaweb.google.com/lh/photo/QWGAgdw27xCo2IjZQmQ0qA?feat=directlink
The picks are not so good because I had both hands on the probes
My "assistant" is an excellent (and I mean excellent) cook but a poor photographer... WS and DATA from non-inverting DAC:
http://picasaweb.google.com/lh/photo/Byy5T1Q8p9oJH9Unycz48w?feat=directlink
WS and DATA from inverting DAC:
http://picasaweb.google.com/lh/photo/PiMieIoOk2bq3gCC68xkSg?feat=directlink
Input signal from parallel DACs to differential opamp, left channel:
http://picasaweb.google.com/lh/photo/v-_UyMzDMTCyduKwAniZRQ?feat=directlink
The same, right channel:
http://picasaweb.google.com/lh/photo/K-bqf5sv76zRZhWwwY4V9Q?feat=directlink
I hope you like it
Cheers,
M
