Digital oscilloscope Project

[davorin]

fpga based digital oscilloscope

Don't forget that you need around 10 times sample frequency to display a certain waveform accurately...for sinewaves 2 times would be okay...but let's say for capturing video signals or even faster ones 2 times is not enough (o;

 

[pisoiu]

chocbar demon

Don't forget that you need around 10 times sample frequency to display a certain waveform accurately...for sinewaves 2 times would be okay...but let's say for capturing video signals or even faster ones 2 times is not enough (o;

Right, Niquist says double, but practice say as you said. Also, introducing besides "1"s and "0"s the uncertainty state in the representation of the signal and also in the trigger decision mechanisms can offer very interesting triggering methods, especially in the area of detecting randomly errors, which are often the hardest to catch.

/pisoiu

 

[davorin]

digital scope schematic

About triggering...make certain you also use registered outputs for trigger decision since can´t expect all datalines are valid at same time...so the trigger logic could see some intermediate values during sample value transition...described in detail at:

www.chocbar.demon.co.uk/download.html

in document DSO_Doc.pdf

 

[ME]

fpga4u.com

Don't forget that you need around 10 times sample frequency to display a certain waveform accurately...for sinewaves 2 times would be okay...but let's say for capturing video signals or even faster ones 2 times is not enough (o;

If the sample rate is too low, the width of the logic pulses would be inaccurate.

Let's say you want to measure a 40% duty-cycle clock.
Then it's not enough to make two samples per clock period, then it would just display a 50 % duty-cycle.

You need to sample at least 10 times per clock period, 20 is better.

But the frequency content in such a signal is also much higher than the clock frequency itself.

-------------------------------------------------------------------------------------

About triggering...make certain you also use registered outputs for trigger decision since can´t expect all datalines are valid at same time...so the trigger logic could see some intermediate values during sample value transition...described in detail at:

www.chocbar.demon.co.uk/download.html

in document DSO_Doc.pdf

That's true, that is why BitScope uses 74HC573 latches instead of 74HC244 buffers:
http://www.bitscope.com/design/hardware/pdf/Bs11-3.PDF

If monnoliv sticks with the present design, it would probably be better to
use SN74LVCH16373A: http://focus.ti.com/lit/ds/symlink/sn74lvch16373a.pdf
or SN74LVC16373A: **broken link removed**
instead of SN74LVCH16244A: http://focus.ti.com/lit/ds/symlink/sn74lvch16244a.pdf
You only need to add a Latch Enable signal to sample the data.

 

[ME]

ad8330

@monnoliv

Some suggestions:

1) I think you need to change the input stage sensitivity.
Standard voltage range for an oscilloscope is 14 positions from 1mV/div to 20V/div in (full 8div range: 8mV/8div - 120V/8div) in 1-2-5 sequence:
1mV/div, 2mV/div, 5mV/div, 10mV/div, 20mV/div, 100mV/div, 200mV/div, 500mV/div, 1V/div, 2V/div, 5V/div, 10V/div, 20V/div.

I don't think your design is sensitive enough for small signals, you have to increase amplification.
Wouldn't it be better to change the 1/10 and 1/100 stage at the input to say: 1/2 and 1/20?

2) Why not move C1||S2 (DC/AC switch) to the other side of R1||C3? The the voltages across S2 would be smaller and it would be easier to replace it with a 'Photovoltaic Relay' as suggested earlier in this topic.

3) I think you have to improve the DC-offset circuit added.
Now it is connected directly to +-5V, but I don't think these supplies are stable enough for this purpose.
With noise and variations on the +-5V supplies it would ony make the problem worse than if you had no DC-offset adjustment.
Maybe you should use +-1.24V voltgae references to connect to the trimmer?
Or just go back to the original design - or figure out a third soloution.
IS the design good enough as it is, what do you think, I'm not so sure?

 

[ME]

tlc549 oscilloscope

@monnoliv. I found three more errors in your opto-coupler design:


1) The opto-couplers powered by UIN from the USB side (U15 & U17) should be replaced by 5V opto-couplers, since the USB voltage is 4.35V to 5.25V. The opto-couplers used can only handle 3.6V.
Equivalent types are:
HCPL-060L (3V, single port) - HCPL-0600 (5V, single port)
HCPL-063L (3V, dual port) - HCPL-0630 (5V, dual port)

The USB chip has a 3.3V LDO to used for the I/O pins to interface with 3.3V logic, but it can only supply 5mA to external components, which isn't enough to supply the opto-couplers.
Another soloution is to place an external 3.3V LDO to supply the low voltage opto-couplers and the USB I/O ports. This might save some current.
The total amount of current whcih could be drawn from the USB port is 500mA. I don't know if it is enough to supply the whole board?

--------------------------------------------------------------------------

2) You need two more lines for the JTAG-interface, one input and one output.
This means that U16 and U17 should be replaced with dual opto couplers.
Check this great documet for Morph-IC: www.morph-ic.com/documents/DSMorphic10.pdf
This is a very exciting document because it is a eval board from FTDI with the same FPGA (Altera ACEX) as used in your design.
Link to eval board site: http://www.morph-ic.com
It shows how to program the FPGA through the USBchip.

Software and source code to the PC for configuring the FPGA is included in the eval kit.

Morph-IC 1K USB Schematic: www.unitronic.de/ftdi/morphic_schematic.pdf
Please note that there is a mistake in this schematic - SI/WUA and SIWUB
should be connected to VCCD!
http://www.unitronic.de

--------------------------------------------------------------------------

3) The LED forward current is too low:
Diode forward voltage max: 1.75
Vcc: 3.3V
Min. forward current 5mA (The initial switching threshold is 5 mA or less. It is recommended that 6.3 mA to 10 mA be used
for best performance and to permit at least a 20% LED degradation guardband.):
**broken link removed**
The pull-up resistor values should just be below 4kohm, so you can replace the 3.3kohm resistors with 3.9kohm to save a little power.

Calculation of the LED resistor value (Vout low of FPGA / USB-IC is assumed to be 0 V, in pratice it is higher, which results in a lower forward current):

3.3V-1.75=1.55V

1.55V / 6.3mA = 246 ohm

The resistor should be lower than 246 ohm.

Use a 220 ohm.

 

[monnoliv]

project osciloscop low frekvency

Hi all,

Many logic analysers have the posibility to choose the type of logic levels sensed at input, so maybe comparators with adjustable levels for Vihmin and Vilmax are required. For example you won't be able to see PECL levels with a TTL buffer at input. Of course, this is if you want this feature installed...

The problem is: what consideration about logic analyser? Is it only a +, then a simple one is good enough, or do we want a complete one (more expensive). If we put 32 fast comparators it will be more expensive (+ bigger PCB + DAC for threshold voltage + ref voltage,...). The question is open...

Negative ECL 10k and 100k logic could also be detected if you use +-5V supply for the comparators and use special open collector comparators to avoid negative levels at the output.

like the AD8561 (I used it in previous revision for triggering) but it's expensive: 4.23*32 > 120 € (4.23 is the price from RS components, they are expensive, I know but it's just an idea).

If monnoliv sticks with the present design, it would probably be better to
use SN74LVCH16373A: https://focus.ti.com/lit/ds/symlink/sn74lvch16373a.pdf

Good idea, I'll make the change.

I don't think your design is sensitive enough for small signals, you have to increase amplification.
Wouldn't it be better to change the 1/10 and 1/100 stage at the input to say: 1/2 and 1/20?

Ok, I'll check.

2) Why not move C1||S2 (DC/AC switch) to the other side of R1||C3? The the voltages across S2 would be smaller and it would be easier to replace it with a 'Photovoltaic Relay' as suggested earlier in this topic.

I've to check also but a photomos is difficult to buy for me (distributors? I don't want to buy 1000 units!)

Maybe you should use +-1.24V voltgae references to connect to the trimmer?

Right, I can check also to have a DAC with integrated ref as a DC-offset adjustment (I think that the price for a votage ref is just less expensive than a DAC with ref but I've to check).

1) The opto-couplers powered by UIN from the USB side (U15 & U17) should be replaced by 5V opto-couplers, since the USB voltage is 4.35V to 5.25V. The opto-couplers used can only handle 3.6V.
Equivalent types are:
HCPL-060L (3V, single port) - HCPL-0600 (5V, single port)
HCPL-063L (3V, dual port) - HCPL-0630 (5V, dual port)

The datasheet isn't accurate for the 5V voltage. It's written 7V absolute maximum voltage for 1 min. I was presuming that 5V was possible... But ok, I'll check to use 5V version.

2) You need two more lines for the JTAG-interface, one input and one output.

I don't think. The morph-IC (it's amazing the similarity in the schematics!!!) is using the passive serial configuration, not the JTAG. Passive serial conf. is a proprietary configuration protocol. The advantage of using JTAG is that you will be able to program also the µC for the hand-held version (if it support JTAG, of course). To see the difference between passive serial conf. and JTAG, here is a doc: https://www.altera.com/literature/hb/cfg/cfg_cf51005.pdf

3) The LED forward current is too low:

I know but the opto-isolator current consumption is high and I wanted to derate a bit ! (anyway, we can change resistor after).

Thank you very much for your comments. The schematics will be updated soon.

Still some post and we'll have a vgood scope (I think the best home-made one :wink: )

Regards,

 

[ME]

ic adc for oscilloscope

1) The opto-couplers powered by UIN from the USB side (U15 & U17) should be replaced by 5V opto-couplers, since the USB voltage is 4.35V to 5.25V. The opto-couplers used can only handle 3.6V.
Equivalent types are:
HCPL-060L (3V, single port) - HCPL-0600 (5V, single port)
HCPL-063L (3V, dual port) - HCPL-0630 (5V, dual port)

The datasheet isn't accurate for the 5V voltage. It's written 7V absolute maximum voltage for 1 min. I was presuming that 5V was possible... But ok, I'll check to use 5V version.

Actually it is says it very clearly in the datasheet, but it is well hidden down at page 6:

Power Supply Voltage Vcc: 2.7 Vmin - 3.6 Vmax

Look at the last table at page 6 called 'Recommended Operating Conditions':
**broken link removed**

2) Why not move C1||S2 (DC/AC switch) to the other side of R1||C3? The the voltages across S2 would be smaller and it would be easier to replace it with a 'Photovoltaic Relay' as suggested earlier in this topic.

I've to check also but a photomos is difficult to buy for me (distributors? I don't want to buy 1000 units!)

I just went to my local electronics store yesterday to collect a new catalouge called Semiconductors.
I opened it and on the 8th page under 'New Products' was three Photovoltaic Relays in 6-pin DIP housing from International Rectifier (www.irf.com):
PVG612 (60V 2A)
PVT312 (250V 0.32A)
PVT412 (400V 0.21A)

Monday I will try to ask for prices and if there is any minimum quantity.
Maybe I can mail some to you, we could start to exchange components, I have a hard time finding an Altera ACEX to a reasonable price.
Xilinx seems more popular in my country, but still expensive for private buyers.

I just checked out International Rectifier's web site, and it turns out they sell them online in single units!
You can also buy as a private person and use credit card.
Maybe you can buy some to sell together with the other special components required monnoliv?
The price for PVT412 is $2.55 per unit.
Subtotal: $2.55
Handling: $5.00
Shipping: $18.00 international
Processing Fee: $6.00

Of course the costs is too expensive if you order a single unit, so you have to buy a certain amount to reduce the shipping and handling costs per unit.


You can also buy International Rectifier's Photo Voltiac Relays at www.farnell.com
Farnell are also present in your country, Belgium.
Just search for 'RELAY PHOTOVOLTAIC' at Farnell and you will find ~37 different devices.
The prices are ~2-3€ per unit and the shipping is cheaper than from International Rectifier.
You can order them together with many of the other parts for the project at Farnell.
Farnell also sells to private persons - at least they do in my country.
The prices are about the same as at RS Components, but it is easier to order as a private person.
They also sell Altera ACEX EP1K30TC144-3 for € 31,36 at Farnell - rather expensive.

3) The LED forward current is too low:

I know but the opto-isolator current consumption is high and I wanted to derate a bit ! (anyway, we can change resistor after).

Yes we can always change the current.
The opto-gates are specified for 15 MBs (Mega Baud pes second I guess it means).
They are going to run a little slower in this design so the LED current can probaby be lowered a little too.
I looked at the specs for some of the slower opto-gates. I was only 0.5mA for a 100kBs device I think.
You could change the values of the pull-up resistors to 3.9kohm to same a little power, it is still under the 4kohm reccomended max value of the pull-ups.
All the JTAG outputs of the USB-IC could probably be turned off/ pulled high by the PC-software after programming has finished.
If only one of the USB ports A and B was drawing current through the LEDs at the same time we, would save some power.

 

[monnoliv]

hobby oscilloscope project

Look at the last table at page 6 called 'Recommended Operating Conditions':

Ok thanks

I just went to my local electronics store yesterday to collect a new catalouge called Semiconductors.
I opened it and on the 8th page under 'New Products' was three Photovoltaic Relays in 6-pin DIP housing from International Rectifier (www.irf.com):
PVG612 (60V 2A)
PVT312 (250V 0.32A)
PVT412 (400V 0.21A) ...... You can also buy International Rectifier's Photo Voltiac Relays at www.farnell.com
Farnell are also present in your country, Belgium.
Just search for 'RELAY PHOTOVOLTAIC' at Farnell and ......

Great, I'll make the upgrade

Maybe you can buy some to sell together with the other special components required monnoliv?

No problem,

Thanks,

 

[maneco]

lmx9838 availability in chennai

Hi,
regarding oscilloscopes,the simplest project seems to be the mk1 or mk3 mentioned preeviously,but i'm not able to find fifo ram in Uruguay...
Does anyone have the MK2 version that appeared in EA some years ago?

Thanks

 

[divan]

digital storage oscilloscope project schematic

RF RELAY OMRON
G6k-RF

Features
# Surface mounting sub-miniature DPDT relay
# Dimensions of 10.3mm x 6.9mm x 5.4mm
# Superior high frequency characteristics of 1GHz


Technical Specifications
Contact Form DPST
Resistive Load 0.3 A at 125 VAC, 1 A at 30 VDC
Max Switching Power 1W
Rated Voltage 3 to 24 VDC
Power Consumption 100 mW
Mechanical Endurance 50,000,000 operations min
Electrical Endurance 100,000 operations minimum
Operating temperature -40° to 70°

 

[martingn]

forum recommend cheap used digital oscilloscope

Hi all,

Monnoliv,
Please find attached one possible solution to enhance the input scale, just figures, no schematic. (just one file xls)

It implies one 4 output divider, one (or two) amplifier and a new gain to the differential amp. May be we can consider another diff with external gain set and lower noise. (AD8008 650mhz dual low noise samecost).

I've a copy of altium DXP so if you need a hand with sch/pcb let me know.

best regards,
MArtin

 

[ME]

gameboy dso pdf

There is no attachment!

We are now considering AD8330, AD8331 or AD8332, check it out.

 

[martingn]

designe of a 5mhz digital oscilloscope

again...

sorry

regards,

MArtin

 

[monnoliv]

schematic diagram of a oscilloscope

Thank you very much martin,

As ME wrote, we are considering using the AD833X family. Thanks for your sheet, I'll check.

Schematics will be updated in one or two days with all corrections/updates.

Regards,

 

[monnoliv]

project-o-scope

Here is the table from martingn updated for the next schematics (not really fixed yet but ok...)

 

[ME]

oscilloscope storage project

Maybe we can consider another diff with external gain set and lower noise. (AD8008 650mhz dual low noise samecost).

AD8008 has no differential output.

 

[martingn]

avr осциллоскоп rs232

Hi all,

yes... ad8008 is not a diff, sorry

Martin

 

[martingn]

bitscope pretrigger

Hi all,

Monnoliv,ME,
Does the Amplifier VGA column means the Diff Gain?
In that case, can the AD8330 have gain=200 @ 100 mhz?
I understand that AD8330 has 150mhz BW(-3db).

Am I right?

regards,

Martin

 

[ME]

oscilloscope pc schema tda8

Hi all,

Monnoliv,ME,
Does the Amplifier VGA column means the Diff Gain?
In that case, can the AD8330 have gain=200 @ 100 mhz?
I understand that AD8330 has 150mhz BW(-3db).

Am I right?

regards,

Martin

Yes VGA means Variable Gain Amplifier, it is the AD8330.
It has 0dB to 50dB (1V/V to 316V/V) adjustable gain @ DC - 150MHz

AD8330 Low Cost DC-150 MHz Variable Gain Amplifier:
www.analog.com/Analog_Root/productPage/productHome/0,2121,AD8330,00.html