next topic »

[on:]

I purchased a USBasp from China.  Probably the first error.  It was inexpensive and also apparently cheap. 

I as inspecting the board to verify the jumpers did what was expected.  A couple pins which where obviously not connected to ground, could have be connected to Tx and Rx.  So I also wanted to check that out.  Looking the board over and comparing it to the schematic at http://www.fischl.de/usbasp/ I found that the USB port was not correctly designed.  I have attached a photo, the Right design and the Rong design.

The question is, how bad a mistake is this?  Will this fry the processor or is there enough current limiting on-board to not worry about?

I could remove the resisters and wire in some small replacements in the correct place, but is it worth it?

[Reply #1 on:]

Won't the wrong design just keep negative voltages from going into the resistors, vs keep them from going into the connector?

[Reply #2 on:]

Won't the wrong design just keep negative voltages from going into the resistors, vs keep them from going into the connector?

I think the primary thing that the zeners will do is clamp the 2 ATMega output pins to 3.3v.  So while the ATMega is trying to make the pins 5v the zener is shunting the power.
I assume that the processor can not source enough current to blow the zeners and I suspect that processor can tolerate the pins being pulled down a bit.

But these are all assumptions and suspicions.   I'm sure that it doesn't help having the input and output running through dropping resistors, but they're small.

[Reply #3 on:]

No Both drawings are accurate, one (the first one) is a 5V bus adaption and the second one is a 3V3 buss. To use it on a 5V bus add the resistors from the first drawing or some just like them. Neither is wrong, just for different bus voltages and the first one shows how to convert the second.
Helpful?

Doc

[Reply #4 on:]

No Both drawings are accurate, one (the first one) is a 5V bus adaption and the second one is a 3V3 buss. To use it on a 5V bus add the resistors from the first drawing or some just like them. Neither is wrong, just for different bus voltages and the first one shows how to convert the second.
Helpful?

I don't follow you.  What do you mean as 3v3 buss?  The supply coming from the USB connector is 5v.  The voltage is going to the ATMega chip.  It is going to output digital logic levels of 0v and 5v.  In the second drawing, when it tries to output a 5v signal, the zener is going to drag it down to 3v3 without the benefit of a current limiting resistor.  Also the output signal going to the USB data lines will be diminished by the inline resistors.  If we assume that the normal USB data line is a high impedance input then maybe the resistors (62 ohms) will have little effect. 

Is it clear why I am confuses?

Can I conceive of a reason the the second circuit could be correct? 
If the input 5v were reduced somewhere to 3v3 before powering the ATMega.  This didn't happen
If the USB data lines were driven at voltages above 3v3.  I don't know that they are.  I am no expert on the USB standard.
If you wanted to protect you ATMega from the high voltage USB data, then maybe the second circuit could be correct.

Basically it does not adhere to the design of Thomas Fischl.

[Reply #5 on:]

I said that the first drawing was ok for 5V and the second for 3V3. The second will obviously draw a lot of current as 3V3 < 5V, nothing more. I wasn't intending to offer you advice about the application only the way the drawings looked to me. Give me about 10 mins and I will re-read the whole thread and either apologize for my possible abruptness or re advise you again.

Doc

[Reply #6 on:]

Both are 3V3 drawings, really. The 68R resistors are protection for noise, and over current. The one with the resistors before the diodes is probably a little 'safer' as it will draw 25 ma @ 1.7V (5V - 3V3 = 1V7/68R = .025A or 25 ma from a 5V bus, if the bus (the 2 wires int?? and the other one) were connected to a 5V bus it would be an issue, likely fry the zeners. That would depend on how much current the device connected could source. However if the device is connected to anything @ 5V that could source current there is an issue. I realize you are having issues with the word 'bus' but don't, it is just a common referral for me i.e. what it is connected to, a 5V or 3V3 processor, I have an interest too, I bought a rip off of a "black Asp" myself, was $3 & change, couldn't resist it as the shipping was free too, what could I loose except some time. Besides it kinda 'looked' good Too.
I am very sorry If in this process I offended you but for me it was really kind of a non issue. It is clearly a 3V3 device and it really needs to talk to a 3V3 device, might work @ 5V but I really don't know. I would recommend however to put the resistors  in front or between the outputs and the zeners. Both board schematics appear to be 3V3 drawings. I haven't investigated farther, even to my 'programmer'. If you like I will try tomorrow again as it is now 11:15 or there about and I am tired. likely making mistakes..

Doc

[Reply #7 on:]

I bought a cheap usb asp of ebay(4$ total), an older version actual of the official design(missing a few jumpers) and I've had no problem uploading and such with it(5v)
im sure they wouldn't sell you something that'll blow up your arduino(I hope), but I haven't seen any negative effects yet

[Reply #8 on:]

Thank you very much, now for a 10 to 6 pin adapter

Doc

[Reply #9 on:]

... im sure they wouldn't sell you something that'll blow up your arduino ...

I'm a little less sure but, I have not seen any problems with the ISP interface.  I don't think it would or could damage the Arduino.  My concern was it damaging itself.

I think to make myself feel better I will probably remove the zeners and replace them in the correct location.  Maybe I'm worried about nothing and there is a lot of over design in some of these parts.

[Reply #10 on:]

... now for a 10 to 6 pin adapter

They too can be had off ebay for $2.50 or so.  It's odd that the adapter with no smarts and almost components costs almost as much as the programmer with a processor, crystal, LEDs and such.

[Reply #11 on:]

Not odd, It's Greed.
Did you see the set of 8 X 8 LED Cube Bare PCB's being offered by that guy in Holland?. Twenty Seven US Dollars for 2 White 7 or 8 in square boards? $120.00 for the whole Kit of Parts?
That I think is Greed or the guy is being robbed himself. These were Kits he was selling. Very Good Looking though. In a month or so Everyone on Ebay selling Electronics will be selling them. A PCB like that from Shenzhen would COST about $0.10 there or so I am told be a friend who lives there. I Bought a package of 805 resistors and one of capacitors (Because they fit well between the .1" Pins used on PCB's) he sent me two of each and a set of inductors (X2) as well. I asked him about PCB's and he just said to send him the art work and he would have a panel done. He also refused to quote prices and made a comment about me not knowing about that city and it's electronics business... Terryking really set me straight on that issue, the guy in china was for real. By definition a panel is the smallest size board that can be 'drilled' with a CNC Drill. When I was doing PCB's here in this country it was about 10 X 10 inches. Sometime it is hard to realize just how cheap everything really is there.

Doc

[Reply #12 on:]

Not odd, It's Greed.

I like greed.  It is a great motivator for someone to work hard and work smart. 
Some believe that greed can lead to theft.  I don't think so.  Sloth when combined with dishonesty leads to theft.  Theft is always accompanied by dishonesty and frequently accompanied by sloth.

Greed has raised more people out of poverty than love, hate or compassion.