Scope manufacturers provide probes that have a BNC shunt capacitance that couples appropriately with the scope shunt capacitance - not all probes will compensate with all scopes, as scopes can have different shunt capacitance specs.
The x100 probe may still work fine, although it may need its BNC shunt capacitor reduced in value or removed.
The x100 probe may still work fine, although it may need its BNC shunt capacitor reduced in value or removed.
Disaster struck today with my new dual channel version.
1/ A pin I used to select voltage on channel 2 is actually used by USB system so I cant use it for voltage selection. I have had to cut the track and connect to a different pin.
2/ A strange fault I found where if the input is too high it crashed out the USB bus.
Turned out I had connected the op-amp to 5 volts instead of 3v3 causing the op-amp to over drive the 4051 and cause high current flow. The fix was cut the op-amp b+ and connect it to 3v3.
A case of rushing the pcb when I should have checked the pin functioned first OK.
Same problem with connecting op-amp to 5 volts instead of 3v3.
1/ A pin I used to select voltage on channel 2 is actually used by USB system so I cant use it for voltage selection. I have had to cut the track and connect to a different pin.
2/ A strange fault I found where if the input is too high it crashed out the USB bus.
Turned out I had connected the op-amp to 5 volts instead of 3v3 causing the op-amp to over drive the 4051 and cause high current flow. The fix was cut the op-amp b+ and connect it to 3v3.
A case of rushing the pcb when I should have checked the pin functioned first OK.
Same problem with connecting op-amp to 5 volts instead of 3v3.
On to the next version now.
I have started a project using a PIC32MZ1024EFG064-I/PT
It runs at 200MHz and the ADC runs up to 28 MIPS.
I have the pcb's in but tried soldering on PIC TQFP 64 pin package and it was a disaster.
Put far too much solder paste on it then heated it up with hot air gun and the solder just ended up in a big blob !
I wicked most of it off but then realised the PIC had moved and was one pad out vertically.
So I removed the PIC and underneath was another blob of solder so it would never have worked.
I have ordered some syringes for the solder paste with very fine needles to apply much less solder paste. Just waiting for them to arrive.
I have started a project using a PIC32MZ1024EFG064-I/PT
It runs at 200MHz and the ADC runs up to 28 MIPS.
I have the pcb's in but tried soldering on PIC TQFP 64 pin package and it was a disaster.
Put far too much solder paste on it then heated it up with hot air gun and the solder just ended up in a big blob !
I wicked most of it off but then realised the PIC had moved and was one pad out vertically.
So I removed the PIC and underneath was another blob of solder so it would never have worked.
I have ordered some syringes for the solder paste with very fine needles to apply much less solder paste. Just waiting for them to arrive.
Last edited:
I found out my pcb for the PIC micro was out a little and so the pads didn't line up accumulatively. By the time it got to the 16th pad it was out by half a pad.
Looked up the data sheet for it and designed a new pcb which is spot on.
This time I printed it out to check for fit before ordering the pcb.
PCB's arrived and I had another go at soldering the PIC TQFP package.
The pins are on 1mm grid ! great fun.
I soldered pin 1 to the pcb then checked with a magnifying glass it was spot on.
With a little bit of nudging I eventually got it bang on.
I then used a SMD solder syringe to put a little solder on the pins.
I then used my iron to melt the solder.
Ended up with four big blobs again !
I managed to wick off most of the solder and then checked with my multimeter there were no shorts. I found a couple but they came off with a light touch of the soldering iron.
I used to the software from a previous project for the USB interface.
Amazingly the USB bus logged onto my pc with no problems.
I found I was getting error when I tried to read the PIC from the PC.
I think there is a problem with the compiler somewhere as this software previous worked although on pic32mx not a pic32mz.
I managed to bodge the software to work and it runs at about 2 mega samples per second. A big change was the a2d converter which is very different from previous PICs I have worked on. A kind gent on another forum gave me some basic code to use and that worked ok if a little slow. I managed to tweak the a2d converters timings up to 2 mega samples per second which isn't too bad.
Looked up the data sheet for it and designed a new pcb which is spot on.
This time I printed it out to check for fit before ordering the pcb.
PCB's arrived and I had another go at soldering the PIC TQFP package.
The pins are on 1mm grid ! great fun.
I soldered pin 1 to the pcb then checked with a magnifying glass it was spot on.
With a little bit of nudging I eventually got it bang on.
I then used a SMD solder syringe to put a little solder on the pins.
I then used my iron to melt the solder.
Ended up with four big blobs again !
I managed to wick off most of the solder and then checked with my multimeter there were no shorts. I found a couple but they came off with a light touch of the soldering iron.
I used to the software from a previous project for the USB interface.
Amazingly the USB bus logged onto my pc with no problems.
I found I was getting error when I tried to read the PIC from the PC.
I think there is a problem with the compiler somewhere as this software previous worked although on pic32mx not a pic32mz.
I managed to bodge the software to work and it runs at about 2 mega samples per second. A big change was the a2d converter which is very different from previous PICs I have worked on. A kind gent on another forum gave me some basic code to use and that worked ok if a little slow. I managed to tweak the a2d converters timings up to 2 mega samples per second which isn't too bad.
I forgot to add the knack with using solder wick is to use plenty of flux then just press the soldering iron against it and move the iron along the wick.
You will see solder coming through the wick.
Don't try to move the wick along the pins or you can end up bending pins if you are too rough.
You will see solder coming through the wick.
Don't try to move the wick along the pins or you can end up bending pins if you are too rough.
I forgot to add the knack with using solder wick is to use plenty of flux then just press the soldering iron against it and move the iron along the wick.
You will see solder coming through the wick.
Don't try to move the wick along the pins or you can end up bending pins if you are too rough.
Started on another PIC32MZ adaptor pcb.
Couldn't get the decoupling and oscillator caps to solder.
Turned out my iron tip was shot.
I had been a little naughty and left the soldering on while I was writing code for hours and this had deteriorated the tip.
A change of tip and the job was much better.
100_0082 | KODAK Digital Still Camera | harrabylad | Flickr
Last edited:
Came across a software problem.
The pc and pic software run great and for as long as I like.
However if I exit the pc program after 5 seconds the pic goes into suspend and never comes out.
This doesn't happen with previous scopes I have designed.
My previous pic32mx runs fine but the newer pic32mz doesn't.
The pc and pic software run great and for as long as I like.
However if I exit the pc program after 5 seconds the pic goes into suspend and never comes out.
This doesn't happen with previous scopes I have designed.
My previous pic32mx runs fine but the newer pic32mz doesn't.
Built up my second PIC32MZ to DIL converter pcb.
Had usual SMD troubles with alignment and shorted pins.
Lots of copper wick and patience and got most of the shorts unshorted.
I left a couple shorted as the pins weren't used anyway.
A bit naughty but when I tested the main pcb and daughterboard it worked fine.
Had usual SMD troubles with alignment and shorted pins.
Lots of copper wick and patience and got most of the shorts unshorted.
I left a couple shorted as the pins weren't used anyway.
A bit naughty but when I tested the main pcb and daughterboard it worked fine.
I found a problem with the USB transmit buffer being corrupted.
It turned out to be that the USB system DMA's the data from the transmit buffer to the USB hardware.
The transmit buffer is cached so sometimes isn't up to date causing wrong data to be sent !
I set the transmit buffer to be "coherent" and now it works fine.
It turned out to be that the USB system DMA's the data from the transmit buffer to the USB hardware.
The transmit buffer is cached so sometimes isn't up to date causing wrong data to be sent !
I set the transmit buffer to be "coherent" and now it works fine.
That setting was for the USB scope (i.e. for transfer over the USB bus ) or for the USB controller (i.e. for transfer for the PCI(e)/DMA/CPU)?
The problem was with the PIC microcontroller.
The DMA ignores cached data so corrupts what is in the transmit buffer.
Working on a new version.
Its a 20 mega samples/second unit.
Uses a AD9200 and PIC32MZ processor to get the speed.
I don't fancy soldering the two small SMD devices so might get the prototypes built in China
Cool to follow your progress. Any relevant shots or pics you can share?
Cool to follow your progress. Any relevant shots or pics you can share?
Its early days for the 20MHz sampling version, still working on schematic at the moment.
While waiting for my new version to come through I built up another revision 16 pcb.
This one is 2 meg samples/second and 200KHz bandwidth.
This one is 2 meg samples/second and 200KHz bandwidth.
An externally hosted image should be here but it was not working when we last tested it.
- Home
- Design & Build
- Equipment & Tools
- PC USB scope project.