Re: [avr-chat] my USBasp programmer does not work but the one from ebay is working

[David Gustavson]

Actually, inductance of the wires is also a potential problem.

With today's fast circuitry, wires should be kept as short as possible. 

For some circuits, even using those breadboards is impractical.

Even 25 years ago I recall working with a circuit that would fail before
one wire got up to one inch (but that was at 250MHz).

The breadboard technique then was to mount the ICs upside down on solid
copper circuit board, and run wires as directly as possible and keeping
them as close as possible to the copper. Capacitors had to have
especially short leads. But for really fast chips you had to have
multilayer boards with narrow conductors carefully spaced in buried
layers.

So, long convenient wires are really asking for trouble. Wires are
inductors with capacitors between them and everything else, as well as
being resistors. They are components.

Dave


On Sun, May 24, 2015, at 10:33 AM, YuGiOhJCJ Mailing-List wrote:
> Really interesting! So, there is a relationship between my problem and
> the length/number of wires.
> I have a fixed length of all my wires: about 8cm (this value corresponds
> to half of my breadboard width).
> The advantage of this length is that I can reach components that are far
> and components that are near (so it is more reusable).
> 
> That is a big constraint to work at 12 Mhz if it requires to have a
> limited length/number of wires.
> I will try to reduce the number of wires and if it does not work, I will
> not have the choice: I will create shorter wires (specific for this
> circuit, so less reusable).
> 
> I have just successfully done a circuit using a lot of useless wires
> here:
> http://yugiohjcj.free.fr/elec-vusb-led-1.jpg
> http://yugiohjcj.free.fr/elec-vusb-led-2.jpg
> It is an USB connected breadboard that is controlled by a program on the
> computer [1, 2].
> I can switch on or switch off the LED.
> In this case, I have no problem... Do you know why I have problems with
> the USBasp project and not with this one?
> 
> [1] https://www.youtube.com/watch?v=iA15vqa0Rk8
> [2] http://www.wzona.info/2012/10/usb-duomenu-perdavimas-atmega.html
> > Hi,
> > The real problem is capacitance on USB wires. Long story short: the USB
> > cables should be short as possible since USB working at 12 MHz (at least
> > USBasp).
> > Also 12 MHz crystal should be very close to AVR chip. Long wires make
> > troubles :)
> > Regards
> >    Martin
> > 
> > > > This is a photo of my breadboard:
> > > > http://yugiohjcj.free.fr/usbasp.jpg
> > 
> > It looks as if you have "too many wires". This schematic should have
> > just the CPU, a couple of wires to USB and a few (4) wires going to
> > the target. Together with two zeners and a two resistors, I expect
> > about 12 wires in total. I'm seeing about 3 times too many.
> > 
> >         Roger.
> >
> 
> _______________________________________________
> AVR-chat mailing list
> address@hidden
> https://lists.nongnu.org/mailman/listinfo/avr-chat