Hi,

 I can't contact with Anita Van Loon about DSR4 chip's price and delivery.

Did you still sell it? Can you provide me this information?

I too have air conditioning in the room, not sure to which power line it is connected, but I'll try to check this aswell.

Thank you very much.

At hight rates I worked with files of up to 20 GB so I don't think this is the problem.

I will try to run it under Ubuntu and see if i can recreate the problem.

Thank you very much for the quick responses and help.

Hi I have also positron annihilation system based on DRS4v4 cards. Its running several weeks, sometimes months, without freezing in windows7 64bit system (Pentium Core Quad 2Ghz, 4GRam).  Problem was when widows was trying to install updates and restarted PC. In beginning I had some problem with memory leak in my application, but it was simple seen in task manager that application memory was rising and was need to find memory leak in application code. Now I remember card was sometimes freezing when room air conditioning with 2kW was starting and high electricity pulses were reason of USB problems, it helped to put air conditioner and PC in different power line input. Hope it help to solve zour problems.

Martin

Even with writing for one night no problem (see below). Have you checked how big your data file is? I guess there is a limit under Windows of 2 GB. If that's the case, you have to write shorter files.

I tried this night to run the board at a 10 Hz rate with an external pulser, without writing, and it did not freeze after ~14 hours of running on Mac OSX. This night I will try again with writing.

Stefan

Then it seems that there is some USB communication problem. I heard this also from other people, that the USB data transfer under Windows has sometimes problems. I develop and run the board under Mac OSX, and there the same software runs for days without problem. So I guess it's related to the underlying libusb lib which is used by the DRS oscilloscope, on which I have no influence. So the only advice I can give is to take shorter series of data. Anyhow the board is not considered a full DAQ system, just an "evaluation board" which means one can try the DRS4 chip and play with it. For serious business one should build own electronics with the chip. Anyhow we are currently developping an Ethernet board which allows much faster acquisition rates, so USB will be obsolete some day. Nevertheless I will try to reproduce your problem and see if I can do anything. At what trigger rate does it show up most prominently?

Stefan

Dear Stefan Ritt,

Yes I use Windows 7, If the DRS Oscilloscope program stays on for a couple of hours without saving the data, the problem will occur. It seems it happens more often when there is data writing and when the rate of events is slow, about 100 events per second, at high rates it almost doesn't happen. Can it be temperature related?

Daniel

Dear Daniel,

sorry my late reply, I'm pretty busy these days. The behavior you report has not been seen before, but I guess no one tried to take such long runs of data yet. Can you confirm that the problem also occurs without writing data to disk, or is it disk-related? I guess you use it under Windows 7, right?

Stefan

Thanks, great!

No there are no other components. I put a photo of the inverter with its cables SMA and one end, BNC at the other. You can see it is very small. I glued the inverter to a piece of thin plywood, and fixed the cables to it before attempting to solder them to the pads on the ferite bead support

Thanks again,   this is very useful,  just another question did you put any other passive elements in the circuit for inverting the signal or just simply swaped the transformer connections? 

The evaluation board normally has 1024 bins per channel. We offer an option with 2048 bins using channel cascading, to capture longer waveform windows. The binary data format is however defined as having 1024 bins. Therefore, for the 2048 bin boards, the software averages over two adjacent cells and saves effectively 1024 bins. The noise of each bin improves this way by sqrt(2). The time however is not very well defined, since you average the voltage of two bins. Therefore, I simple also average over the time of the two bins. Maybe this is not the best way, so feel free to change this.

Stefan

Hi,

I am trying to understand some details about the timing calibration.

We wrote our own code but we more or less use the ideas of the Oscilloscope class.

In the binary file writing of in the function Osci.cpp::SaveWaveforms() (line 924ff)

the following code is executed:

if (m_waveDepth == 2048) {
    t = (tcal[j]+tcal[j+1])/2;
    j++;
} else
    t = tcal[j];

I do not understand the averaging of the to adjacent calibration constants. Could you explain this? Do one have two measurements?

Cheers

Felix

The coilcraft part number is: JA4220-ALB. Iordered two of them and they were sent as free samples. You might want to buy some slightly bigger ones. I found them so small it was very hard to solder the coax cable to the connectors. Since I got them, I managed to damage one as they are quite fragile! In the confirmation email I got there was some contact info which may be useful for you: "For help, contact Victoria Berner at 847-516-5551  vberner@coilcraft.com "  BTW every time I use this forum I'm told that my password is not valid. Each time I reset it according to the "Lost pasword preceedure. Then I can log on again. Its quite annoying.

Hi Chris,

 I am looking at Sensl SiPMs as well,  can you send the part number from Coilcraft?

Thanks!

Abaz

I needed a fast pulse inverter in order to feed signals from the recent SensL SiPMs into a conventional CFD which only accepted negative signals. I got a very small ferite torridal transformer from Coilcraft and wired up to invert signals then inserted into in 50 ohm coax cable and it works fine. These things cost only a few cents each!

Thanks, that looks just fine.

Here is one (SI 100): https://www.picoquant.com/products/category/accessories/adapters-splitters-cables-various-accessories-for-photon-counting-setups

Ok, thanks!  do you know an easy in-line inverter like mini-circuit or digikey?    Can also redesign the detector I gues to produce positive signals, just it might be easier if there was a simple inverter if you are aware of? thanks Abaz