[Ed: A thread in which ztex is blamed for a faulty board]

[ztex]

Here is ztex. I'm producing FPGA Boards, see http://www.ztex.de. I also developed a FPGA board which is optimized for cryptographic applications like bitcoin mining: http://www.ztex.de/usb-fpga-1/usb-fpga-1.15x.e.html

Open Source Bitcoin mining software I released here: http://www.ztex.de/btcminer. I'm writing this because scammer usually invest time in developing Open Source Software for the products.

As Gusti wrote, he returned two boards 2 days ago. Both voltage regulators on both boards where destroyed. These voltage regulators are over current and temperature protected. The only way do destroy voltage regulators is by over voltage.

Since the valid range of 4.5 to 16 V (absolute maximum: 4.5 - 18 V)  is clearly stated on the products page (http://www.ztex.de/usb-fpga-1/usb-fpga-1.15x.e.html#con5), this is not covered by warranty.

Here is an image which shows the 3.3V regulator of one of Gustis board:


IMHO Gustie used an unregulated power supply. These things are called unregulated because their real output voltage is not equal to the value on the label. Gusti, did you measured the mean/RMS  voltage or the peak voltage?

The universal supplies from supermarket I mentioned (these one with switchable output voltage) are AFAIK always regulated. Therefore I recommend them as save.

Gusti, you wrote that smoke came out of one part of the first board. Why you destroyed the second board in the same way instead of asking me what is going wrong?

I believe this is a plain SCAM, so , beware from buying anything from ZTEX.

These regulator circuits are used in 4 other products. I sold a few hundreds of them.  Gusti is the only customer I know who destroyed the regulators in this way.

BTW, I just counted it: I wrote Gusti 22 support mails which was not always easy. His last mail was

no problem, bro
I will make sure you will not sell a single card in the forum anymore

EDIT:

Total price including overseas shipping was USD 929.-, I told [redacted] I was going to pay with bitcoins, as he told me he would be accepting them. Price with bitcoins was 10% more than with USD, I was not very happy with that , but I accepted the surplus to make transaction faster.

The 10% included a volatility margin and exchange + transfer frees. In these days the bitcoin price jumped by several percent within minutes.

If you would have used Paypal you would have received the board within the same time.

[1714933022]

1714933022

[ztex]

Last October 21st I ordered 2 x ZTEX USB-FPGA Module 1.15x from manufacturer, through [...]

Gusti, plaese be son kind and remove my private date.

This forum in not platform for a personal vendetta.

[ztex]

Cisco, pricey but rock solid. Never saw a router blow itself as this ZTEX FPGA board crap.

You are writing the FPGA Board has to support your power supply because your router does support it.

In don't understand this logic.

[ztex]

I know this wont make you feel any better, but those volt regs can be replaced.  In fact, if its the 8A Ezbuck one used on this board I have several hundred of them and would be happy to mail you 2.

I excessively tested the circuits a while ago (on a Experimantal Board 1.3 which has the same power supply circuits): If an over voltage occurs the input voltage breaks through. I.e. the chance that FPGA and microcontroller survives is almost zero.

Replacing the 1.2V regulator is very difficult (and expensive) because bottom pads have to be soldered. It is impossible to do this by hand with lead free solder.

[fizzisist]

IMHO Gustie used an unregulated power supply. These things are called unregulated because their real output voltage is not equal to the value on the label. Gusti, did you measured the mean/RMS  voltage or the peak voltage?

The universal supplies from supermarket I mentioned (these one with switchable output voltage) are AFAIK always regulated. Therefore I recommend them as save.

I don't get it, didn't he say he measured 13.5 V open circuit on the regulator? Unregulated power supplies will always output a lower voltage with a load. So, no significant current at 13.5 V should have been applied. As soon as some current was drawn, the voltage would drop. Why would that have caused this damage?

My guess is that the polarity was wrong, although gusti said he checked that.

I'm very curious about what actually caused the problem, because we decided to include a barrel power connector at the last minute on our board and it makes me a little nervous that customers might pull out some crazy power supplies and plug them in. The 4 pin Molex connector will generally be safer, because the output from an ATX PSU will pretty much always be the same.

As a fellow hardware manufacturer, I sympathize with ztex here, but I hope something can be worked out to at least avoid this type of problem in the future. ztex, did you include any fuses or reverse polarity protection on the power input by any chance?

[ztex]

I don't get it, didn't he say he measured 13.5 V open circuit on the regulator? Unregulated power supplies will always output a lower voltage with a load. So, no significant current at 13.5 V should have been applied. As soon as some current was drawn, the voltage would drop. Why would that have caused this damage?

The boards where destroyed in idle mode, i.e. when the FPGA was not configured.

If the power supply dos not have bulk capacitors the voltage you measure is some kind of a mean voltage, usually the RMS voltage, not the peak voltage.

My guess is that the polarity was wrong, although gusti said he checked that.

This would be another explanation.

I'm very curious about what actually caused the problem, because we decided to include a barrel power connector at the last minute on our board and it makes me a little nervous that customers might pull out some crazy power supplies and plug them in. The 4 pin Molex connector will generally be safer, because the output from an ATX PSU will pretty
much always be the same.

As a fellow hardware manufacturer, I sympathize with ztex here, but I hope something can be worked out to at least avoid this type of problem in the future. ztex, did you include any fuses or reverse polarity protection on the power input by any chance?

As I wrote earlier, I sold several hundreds of this circuits, and Gusti is the only one I know who destroyed the regulators in this way.

A protection against such incidents would be a add a power supply. But if you are shipping all over the world you get a problem: different countries have different plugs and different voltages. Expensive international kits would have to be used and due to the size shipping gets even more expansive. Furthermore, most of these 1.15x FPGA Board's are used in clusters with a special power supply, see initial post of https://bitcointalk.org/index.php?topic=49180.0.

Fuses give no additional protection since voltage regulators are over current protected.

Polarity protection by diodes would cost 0.5W on my FPGA boards (voltage drop of 0.65V @ 0.7A).

The USB-FPGA Modules 1.15x have low cost bulk capacitors (C3). The only job of this capacitor is to reduce the peak voltage in unregulated of unregulated power supplies. But this is only a small protection which helps to survive the most unregulated 12V supplies. Another not-100% protection for similar scenarios would be a Z-Diode in parallel. But I'm not sure whether these diodes exist.

A 100% protection is difficult / expensive (requires a voltage monitor circuit) and a transistor that switches of the power supply.

Unfortunately, there exists no protection against customers who destroy the second board in the same way after smoke came out of the first one.

[DiabloD3]

Angry bastard mod here.

I've had multiple people report this thread. I think this is the kind of case where I tell both sides to just hire lawyers and sue, I'm locking the thread.