Re: USB multimeter application released

[Chris Bell (News)]

Dave Higton <da...@davehigton.me.uk> wrote:
> I've released an application to display, and optionally log data
> from, a Maplin N56FU USB-interfaced digital multimeter.

Now that sounds REALLY useful - many thanks, Dave. I'll have to think
about buying a USB-enabled meter!

Chris Bell.

--

[Dave Higton]

In message <534aca4...@highpath.net>

I hope it proves to be so, Chris. Do let me know!

Some file dates suggest I started the app in 2008. It's only since
the "How to do unpredictable USB input" thread in the ROOL General
forum that the information necessary for any decent performance has
been assembled.

Dave

[cfe...@freeremoveuk.com.invalid]

In message <0ca5d04a5...@my.inbox.com>

Nice one Dave H

Anyone here tried a "PC Oscilloscope" ie a USB driven scope using a
computer for display.

There was a RO program called !Scope (Mike Cook) driving a board
connected to the parallel port.



--
Colin Ferris Cornwall UK

[Dave Higton]

In message <7cc2d54a5...@cferris.freeuk.com>

cfe...@freeRemoveuk.com.invalid wrote:

>Anyone here tried a "PC Oscilloscope" ie a USB driven scope using a
>computer for display.

Years ago, I have used a PicoScope; but it was with their own
Windows software.

>There was a RO program called !Scope (Mike Cook) driving a board
>connected to the parallel port.

The difficulty is getting decent performance without racking up a
price of hundreds of pounds. It's all very educational to make a
capture system capable of 1 megasample per second or so, but it's
of no use in the real world. Even 100 Msa/s is restrictive. If
you're restricted to anything like 1 Msa/s, you can't even look
sensibly at parasitic oscillation on an audio power amplifier,
never mind anything faster.

I have next to me, on long term loan from the company I work for,
a digital oscilloscope with 4 channels, capable of sampling rates
up to 100 Msa/s per channels, and with 50 k samples per channel
per capture. The sampling rates are independent of the number of
channels running. This is very good for zooming in to get fine
detail, and with a long capture. But even that is far too slow
to find problems with reflections on PCB traces.

Even with logic families of 20 years ago, we needed a minimum of
400 Msa/s to find those problems. 1 Gsa/s was better.

You need dedicated hardware for capture, and then suck the data
back over USB for display and analysis. The challenge is to get
that hardware at a hobbyist price.

Dave

[Stuart]

In article <534aca4...@highpath.net>,

Chris Bell (News) <ne...@highpath.net> wrote:

First thing I need is a USB enabled computer!

From time to time I have contemplated a Unipod but never could quite
justify the expense. From many posts I saw in these parts, USB always
seemed to be a bit hit and miss.

--
Stuart Winsor

Midlands RISC OS and Raspberry pi show, 13th July 2013

http://www.mug.riscos.org/show13/MUGshow.html

[Theo Markettos]

Dave Higton <da...@davehigton.me.uk> wrote:
> You need dedicated hardware for capture, and then suck the data
> back over USB for display and analysis. The challenge is to get
> that hardware at a hobbyist price.

Once you have that hardware, there's not much saving by not having a UI. A
computer based UI has the disadvantage that you need an extra screen, and
you can't easily twiddle the knobs to hunt for a signal.

I've used scopes which have three sorts of UI together: traditional knobs, a
touchscreen, and run Windows XP so you can use a keyboard and mouse. I
mostly used the knobs. The advantage of having a computer more tightly
coupled is you can get the data off more easily - you can pull out the data
into a processing package much more easily. (I spent quite a while writing
obscure GPIB commands to get this out of non-computerised instruments). Or
if you can do your analysis on the data stream in real time as it comes in.
Most scopes have fairly limited mathematical operations they can perform,
so it's nice to be able to write a function and see it applied on the data
live - handy for hunting around for signals, or for 'production test' (show
good/bad straightaway without post-processing).

These days you can buy a full Android tablet for 40 pounds, so there's not a
great deal of money to be saved there. And the higher in frequency you go
you start spending a lot of money on ADCs and an analogue frontend that tries
to be linear at those frequencies, which you need whatever happens on the
digital side.

Theo

[Jim Lesurf]

In article <9f8edc4a5...@my.inbox.com>, Dave Higton

<da...@davehigton.me.uk> wrote:
> In message <7cc2d54a5...@cferris.freeuk.com>
> cfe...@freeRemoveuk.com.invalid wrote:


> >There was a RO program called !Scope (Mike Cook) driving a board
> >connected to the parallel port.

> The difficulty is getting decent performance without racking up a price
> of hundreds of pounds. It's all very educational to make a capture
> system capable of 1 megasample per second or so, but it's of no use in
> the real world. Even 100 Msa/s is restrictive. If you're restricted to
> anything like 1 Msa/s, you can't even look sensibly at parasitic
> oscillation on an audio power amplifier, never mind anything faster.

While that is true for many purposes, you can find an 'audio' scope useful
as well. I've certainly found using the Iyonix as a 'scope' FFT / data
logger useful at times even though limited to the audio range. And I am
hoping to be able to use the ARMiniX's 96k sample rate input for a
scope/FFT/logger sometime when RO can listen to it!

Personally, for high frequencies I find a spectrum analyser more useful
than a wideband scope. That's why I bought a FUNcube Pro Plus and wrote
some software to use it as an RF spectrum analyser. Works nicely from long
wave up to 2 GHz. Alas, at present I can only do this with Linux as the USB
support for it with RO is absent. As is the support for so many other USB
items like good DACs, microscopes, etc, etc. :-/

I find this annoying as better USB support would help us avoid many of the
historic limitations that tied RO to quite specific target hardware.

Slainte,

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scots_Guide/intro/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

[Dave Higton]

In message <534add2d...@argonet.co.uk>

Stuart <Spa...@argonet.co.uk> wrote:

> From time to time I have contemplated a Unipod but never could quite
> justify the expense. From many posts I saw in these parts, USB always
> seemed to be a bit hit and miss.

I'd suggest that a BeagleBoard xM is not a lot more expensive than
a Unipod, but has more extensive USB facilities, plus a huge speed
boost over a Risc PC. Admittedly that idea of cost is based on a
DIY case and having all the other necessary peripherals already.

It's also the subject of considerable ongoing OS development.

USB is a bit hit and miss everywhere - even Windows isn't perfect.
But what else is there?

Dave

[spampling]

In article <534add2d...@argonet.co.uk>,

Stuart <Spa...@argonet.co.uk> wrote:
> First thing I need is a USB enabled computer!

um, 25 quid plus VAT. 39 quid for a starter kit from CPC. Breaks the
bank...

--

Steve Pampling

[Stuart]

In article <c986574b5...@my.inbox.com>,

Dave Higton <da...@davehigton.me.uk> wrote:
> I'd suggest that a BeagleBoard xM is not a lot more expensive than
> a Unipod, but has more extensive USB facilities, plus a huge speed
> boost over a Risc PC. Admittedly that idea of cost is based on a
> DIY case and having all the other necessary peripherals already.

'Tis a thought.

DIY cases not a problem.

I've been "messing about" with electronics since the 60s and become quite
skilled at fabricating my own enclosures out of sheet Aluminium. These
days I could even just mill a beagleboard sized one out of the solid. (not
CNC though)

[Thomas Milius]

In message <534b1dc...@audiomisc.co.uk>

Jim Lesurf <no...@audiomisc.co.uk> wrote:

> Personally, for high frequencies I find a spectrum analyser more useful
> than a wideband scope. That's why I bought a FUNcube Pro Plus and wrote
> some software to use it as an RF spectrum analyser. Works nicely from long
> wave up to 2 GHz. Alas, at present I can only do this with Linux as the USB
> support for it with RO is absent. As is the support for so many other USB
> items like good DACs, microscopes, etc, etc. :-/

I tried an Agilent DAQ some years ago. It is not a problem of the USB stack
itself. The high end devices are using their own mostly undocumented USB
interface classes. So no chance to develop an own driver for the interfaces
are complex and are including calibration etc. With LINUX it is such a
thing. Some devices are supported by the manufacturer but only by drivers of
the manufacturer which are not open source and may fail on other LINUX
versions.

The only devices which could work are National Instrument., Howver I had no
time until now to look through the MegaBytes of documenation.

Best Regards

Thomas Milius

[spampling]

In article <534b5da4...@argonet.co.uk>, Stuart

<Spa...@argonet.co.uk> wrote:
> I've been "messing about" with electronics since the 60s and become
> quite skilled at fabricating my own enclosures out of sheet Aluminium.
> These days I could even just mill a beagleboard sized one out of the
> solid. (not CNC though)

Carpentry instead - oak, teak, mahogany (plenty of bits of old benches
thrown out that landed in peoples garages)
My use of an LTO3 tape box for the beagle was just a case of proving how
cheap it could be. I have a genuine chassis for it not used yet.

On the cheap theme - rip the innards out of a standard height CD drive
and use the metal shell as the basis for a case.

--

Steve Pampling

[Jim Lesurf]

In article <d4785f4b...@thomas-milius.t-online.de>, Thomas Milius

<Thomas...@t-online.de> wrote:
> In message <534b1dc...@audiomisc.co.uk> Jim Lesurf
> <no...@audiomisc.co.uk> wrote:

> > Personally, for high frequencies I find a spectrum analyser more
> > useful than a wideband scope. That's why I bought a FUNcube Pro Plus
> > and wrote some software to use it as an RF spectrum analyser. Works
> > nicely from long wave up to 2 GHz. Alas, at present I can only do this
> > with Linux as the USB support for it with RO is absent. As is the
> > support for so many other USB items like good DACs, microscopes, etc,
> > etc. :-/

> I tried an Agilent DAQ some years ago. It is not a problem of the USB
> stack itself. The high end devices are using their own mostly
> undocumented USB interface classes.

One of the advantages of the FUNCube is that it follows open standards. So
is easy to get working with systems that support them.

The only snag I know of is that some systems (e.g. the Debian distro for
RPi) lacks some a 'dev' package you need to install. Its a normal package
for x86 distros but it seems no-one as yet has bothered to port it to ARM
and the relevant repositories. I know a few people want to sort this out as
the RPi + FUNCube is an attractive portable compact RF analyser and
logging/measurement system. Far cheaper than the thousands of quid for the
RF kit from the 'usual suspects' like Agilent. And since it provides an IQ
output you can then get the raw data and process that however you want.

When working on testing the FUNCube I used an old HP analyser. This shows
the limitations of a lot of such proprietary kit. Complex to use, and you
would have to use HP (as was) special memory cards. Whereas the smaller and
cheaper FUNCube can be used however you fancy. And delivers comperable
results once carefully calibrated.

If interested you can see some of my calibration results at

http://jcgl.orpheusweb.co.uk/funcube/FUNcalibrate.html

and fig 5 on

http://www.audiomisc.co.uk/RadioAndTV/4G/jamming.html

shows an example of a spectrum produced with it. I've also built a noise
source so I can do gain/loss vs frequency measurements on UHF filters,
amps, etc.

FWIW Having done an RF swept analyser and an FFT analyser for it, I plan
sometime to use it as a high performance FM stereo tuner just to see what
is possible. It's all 'software defined'... 8-]

Similarly, the best quality audio USB DACs all support the same asynch/iso
audio modes. So can 'just work' on Linux/Windows/Macs. No 'driver'.
Although you may have to fiddle about to optimise results.

The details of how to interface these are all open. And a number of devices
now use those standards. The reality looks like this via USB is becoming
the norm. Just look how many USB connections there are *inside* an ARMiniX
box! So if RO were to adopt these standards a lot of kit would become
accessible and bypass a lot of the 'case by case' fixing as different
boards like the RPi and Panda are adopted.

[Theo Markettos]

Jim Lesurf <no...@audiomisc.co.uk> wrote:
> When working on testing the FUNCube I used an old HP analyser. This shows
> the limitations of a lot of such proprietary kit. Complex to use, and you
> would have to use HP (as was) special memory cards. Whereas the smaller and
> cheaper FUNCube can be used however you fancy. And delivers comperable
> results once carefully calibrated.

Hmm, interesting. What kind of noise floor do you get out of it? From one
of the pictures it appears to be about -70dBm?

I've seen 'spectrum analyser' devices using off-the-shelf TV tuner chips, so
this is quite interesting in that it uses a cheap TV decoder (I think) but
puts a purpose-designed analogue frontend in front. Just a pity there's a
big gap in the range (between VHF TV/DAB and UHF TV).

One advantage of the USB form factor is you can stick it up your mast or
wherever the signal is and run USB from there, rather than having to run
lossy coax to your computer.

> FWIW Having done an RF swept analyser and an FFT analyser for it, I plan
> sometime to use it as a high performance FM stereo tuner just to see what
> is possible. It's all 'software defined'... 8-]

Did you try any of the 10 pound TV tuners (mostly RTL2832U based)?

Theo

[Jim Lesurf]

In article <p5f*2p...@news.chiark.greenend.org.uk>, Theo Markettos

<theom...@chiark.greenend.org.uk> wrote:
> Jim Lesurf <no...@audiomisc.co.uk> wrote:
> > When working on testing the FUNCube I used an old HP analyser. This
> > shows the limitations of a lot of such proprietary kit. Complex to
> > use, and you would have to use HP (as was) special memory cards.
> > Whereas the smaller and cheaper FUNCube can be used however you fancy.
> > And delivers comperable results once carefully calibrated.

> Hmm, interesting. What kind of noise floor do you get out of it? From
> one of the pictures it appears to be about -70dBm?

The apparent noise floor varies with the settings. The Fig 5 plot was done
using the LNA IIRC but comes via a loft distribution amplifier. And the
values shown in Fig 5 are relative to the 0dB level that maxes the sample
values with the gain I set. **Not** referenced to a value in dBm. So you
can't easily deduce the FUNCube noise floor from that plot.

I've not yet done a noise floor measurement. To me it seems 'adequate' for
my purposes, so not got around to it yet.

Note also that Fig 5 also has a resolution bandwidth in the plot of 200kHz
so you'd need to think of it as a noise power spectral density. If you FFT
or filter the output the noise per 'frequency bin' will fall accordingly.
The program I wrote and use for this does 8k FFTs, but there are various
other SDR progs about that do similar things.

> I've seen 'spectrum analyser' devices using off-the-shelf TV tuner
> chips, so this is quite interesting in that it uses a cheap TV decoder
> (I think) but puts a purpose-designed analogue frontend in front.

Not quite. It uses a silicon tuner chip for the tuning control, etc. But
has a set of its own decent spec RF amp and band filter stages. It also
uses 16 bit samples after the IQ demod so has a bigger dynamic range than
the TV tuners which all seem to use an 8 bit sampler.

> Just a pity there's a big gap in the range (between VHF TV/DAB and UHF
> TV).

Yes, that doesn't bother me too much as my interest tends to be in the
regions around the VHF, DAB, and UHF TV. And most of the cheap TV tuners
have the same sort of 'hole'. Apart from that it covers about 15 kHz (yes
kHz) up to 2 GHz.

> One advantage of the USB form factor is you can stick it up your mast or
> wherever the signal is and run USB from there, rather than having to run
> lossy coax to your computer.

I think I'd be wary of RX or USB noise for that if you wanted ultimate
sensitivity.

> > FWIW Having done an RF swept analyser and an FFT analyser for it, I
> > plan sometime to use it as a high performance FM stereo tuner just to
> > see what is possible. It's all 'software defined'... 8-]

> Did you try any of the 10 pound TV tuners (mostly RTL2832U based)?

I did buy one meaning to try it. But the results I've had from the FUNCube
are so good I've just gone on using that! :-)

[Jim Lesurf]

On 14 May, no...@audiomisc.co.uk wrote:
> In article <p5f*2p...@news.chiark.greenend.org.uk>, Theo Markettos
> <theom...@chiark.greenend.org.uk> wrote:

> > Hmm, interesting. What kind of noise floor do you get out of it?
> > From one of the pictures it appears to be about -70dBm?

> The apparent noise floor varies with the settings.

This may help a bit...

http://jcgl.orpheusweb.co.uk/temp/nowandthen.png

It is a display of the spectra I got here on two days before and after the
local 'reshuffle' of DTTV to make space for 4G above ch60. Plan to publish
it as part of an article in due course to show the DTTV signal levels, etc.

Note the scale is in dBmV, not dBm. And is per 0.2 MHz bin.

IIRC the conversion for 75 Ohm (which this is calibrated for as it is for
domestic 'TV') is that 0 dBmV corresponds to about -48 dBm. So the apparent
noise floor here is somewhere below -100 dBm per 0.2 MHz.

This value isn't completely reliable for various reasons. The signals are
measured at the output from a UHF distribution amp, so include the noise
from that. The IQ ADCs also have some dc offsets I have to remove by
averaging over the rms and that isn't perfect. But it gives some idea of
the possible performance of the FUNCube.

Note that the FUNCube is made for Radio Amateurs interested in things like
picking up weak signals from satellites. And in fact I know some pro radio
astronomers who are taking to using FUNcubes.

Note also the output is 16 bit but the sampling is actually 24 bit. You can
set a scaling gain for the truncation to get a 90dB range over a bigger
total range because of this. The LO is also very stable but I haven't
measured the phase noise.

[cfe...@freeremoveuk.com.invalid]

In message <534b9d9...@audiomisc.co.uk>

Jim Lesurf <no...@audiomisc.co.uk> wrote:

> In article <d4785f4b...@thomas-milius.t-online.de>, Thomas Milius
> <Thomas...@t-online.de> wrote:
>> In message <534b1dc...@audiomisc.co.uk> Jim Lesurf
>> <no...@audiomisc.co.uk> wrote:

[snip]

> The details of how to interface these are all open. And a number of devices
> now use those standards. The reality looks like this via USB is becoming
> the norm. Just look how many USB connections there are *inside* an ARMiniX
> box! So if RO were to adopt these standards a lot of kit would become
> accessible and bypass a lot of the 'case by case' fixing as different
> boards like the RPi and Panda are adopted.

What info does !USBinfo give on the FUNcube dongle?

[John Rickman Iyonix]

spampling wrote

> In article <534b5da4...@argonet.co.uk>, Stuart
> <Spa...@argonet.co.uk> wrote:
>> I've been "messing about" with electronics since the 60s and become
>> quite skilled at fabricating my own enclosures out of sheet Aluminium.
>> These days I could even just mill a beagleboard sized one out of the
>> solid. (not CNC though)

> Carpentry instead - oak, teak, mahogany (plenty of bits of old benches
> thrown out that landed in peoples garages)

my first pi dwells in a luxury mahogany cabinet:-
http://rickman.orpheusweb.co.uk/nolink/pi.jpg
The ex-binocular prism on top is necessary to see the indicator LEDs
as the mahogany blocks the line of sight you get with the clear cases.

> My use of an LTO3 tape box for the beagle was just a case of proving how
> cheap it could be. I have a genuine chassis for it not used yet.

> On the cheap theme - rip the innards out of a standard height CD drive
> and use the metal shell as the basis for a case.

Not so cheap, but I am tempted to replace the innards of my now
defunct A9Home by another Raspberry Pi



--
John Rickman - http://rickman.orpheusweb.co.uk/lynx
No s� nada, y no estoy seguro de �so

[Stuart]

In article <6d16be4b5...@rickman.argonet.co.uk>,

John Rickman Iyonix <ric...@argonet.co.uk> wrote:

> Not so cheap, but I am tempted to replace the innards of my now
> defunct A9Home by another Raspberry Pi

Oh dear John, how did you kill it?

[Jim Lesurf]

In article <af17b74b5...@cferris.freeuk.com>,

<cfe...@freeRemoveuk.com.invalid> wrote:
> In message <534b9d9...@audiomisc.co.uk> Jim Lesurf
> <no...@audiomisc.co.uk> wrote:

[snip]

> What info does !USBinfo give on the FUNcube dongle?

I've just used !USBinfo both with and without the FUNCube Pro Plus
connected to my ARMiniX. I think this is what the FUNCube adds, but please
let me know if more info is needed. On Linux is presents as both

A) a 'soundcard' for the output as two 16 bits streams of 'stereo audio'
at 192k sample rate which are actually the IQ output from the RF mixers.

B) an HID device for control and settings data. The details of this are in
the QTHID package I used as the basis for my Linux software. I can point at
this more explicitly if useful. The source code for Linux is actually in
the FUNCube apps on my software page on my audiomisc.co.uk site.

Details of !USBinfo output follow...


Device: 12
DeviceFS : USB13
USB spec : 1.10
Class : 0 (Defined at Interface level)
Subclass : 0
Protocol : 0
MaxPacket : 8
Vendor : 04d8 Microchip Technology, Inc.
Product : fb31
Release : 20.03
Speed : Full speed (12 Mb/s)
NumConfigs: 1

Config: 1
MaxPower : 200 mA
Attributes:
Interfaces: 3

Interface: 0
Alternate: 0
Class : 1 Audio
Subclass : 1 Control Device
Protocol : 0 00
Endpoints : 0

> Report=36 Length=9
> 09 24 01 00 01 1e 00 01 01

> Report=36 Length=12
> 0c 24 02 01 03 06 00 02 00 00 00 00

> Report=36 Length=9
> 09 24 03 02 01 01 00 01 00

Interface: 1
Alternate: 0
Class : 1 Audio
Subclass : 2 Streaming
Protocol : 0 00
Endpoints : 0

Interface: 1
Alternate: 1
Class : 1 Audio
Subclass : 2 Streaming
Protocol : 0 00
Endpoints : 1

> Report=36 Length=7
> 07 24 01 02 01 01 00

> Report=36 Length=11
> 0b 24 02 01 02 02 10 01 00 ee 02

Endpoint: 1
Direction : IN
Transfer : Isochronous
Interval : -

> Report=37 Length=7
> 07 25 01 01 00 00 00

Interface: 2
Alternate: 0
Class : 3 Human Interface Devices
Subclass : 0 No Subclass
Protocol : 0 None
Endpoints : 2

HID Descriptors available:
Report (type 0x22, 29 bytes)

Endpoint: 2
Direction : IN
Transfer : Interrupt
Interval : 1 ms

Endpoint: 2
Direction : OUT
Transfer : Interrupt
Interval : 1 ms

Hope that make sense. If not I can post the full report, etc.

Cheers,

[John Rickman Iyonix]

Stuart wrote

> In article <6d16be4b5...@rickman.argonet.co.uk>,
> John Rickman Iyonix <ric...@argonet.co.uk> wrote:

>> Not so cheap, but I am tempted to replace the innards of my now
>> defunct A9Home by another Raspberry Pi

> Oh dear John, how did you kill it?

The hard disk was dodgy from the start, but one day it just gave up
altogether and refused to to spin.

John

--
John Rickman - http://rickman.orpheusweb.co.uk/lynx

"Poetry is a MUG's game" - TS Eliot

[Stuart]

In article <c369294c5...@rickman.argonet.co.uk>,

John Rickman Iyonix <ric...@argonet.co.uk> wrote:
> >> Not so cheap, but I am tempted to replace the innards of my now
> >> defunct A9Home by another Raspberry Pi

> > Oh dear John, how did you kill it?

> The hard disk was dodgy from the start, but one day it just gave up
> altogether and refused to to spin.

If you fancy donating it to a worthy cause - me - bring it along on
Saturday. I'd like to have a play with it