New bench scope - Fnirsi 1014D, 7", 1GSa/s

[pcprogrammer]

Do you know that this sometimes generates errors?

Sometimes is an understatement. I have seen it go wrong quite a lot, almost in every communication attempt it fails only to recover due to repeated tries.

[yangyao]

Do you know that this sometimes generates errors?

Yes, I know, it has been solved in the new version of the machine, the clock frequency of the I2C bus is too high

[yangyao]

@yangyao: A question for you. Can you tell us what the "special IC" in the 1013D, 1014D and possibly the DPOX180H is. I'm talking about the 8 pin chip that is connected to the FPGA based on a more or less I2C hardware protocol, and that "translates" information send from the F1C100s to something else based on a fixed algorithm?

And can you explain the reasoning of why this setup has been used?

Is this chip also used in any other of your designs like for instance the DPOS350P. I can not see if it is there. The 8 pin chip besides the Gowin FPGA might be it or it is the FLASH for the FPGA configuration.

A bit of advice in relation to becoming a more professional test equipment provider, is to make sure that the FPGA configuration can be updated from the MCU. For instance the Hantek DSO2000 series, that also uses a Gowin FPGA, uses the slave serial configuration option the FPGA provides. This means that the bitstream resides in the FLASH memory that also holds the firmware.

That is an encryption chip, which cannot be cracked. I blocked all information interfaces, including physical bus information, ARM IO interface, ARM algorithm, FPGA algorithm, everything is random, in order to prevent other peers from directly copying the product and causing huge losses.

Only when you changed the setup of it after developing the 1013D and 1014D then. Because in those two I cracked it with not to much effort.

If the original software is not needed and the software is newly developed, then the encryption chip is useless. However, in the new product DPOS350P, the encryption chip has been completely separated from the ARM software system, which means that the newly developed software must also be developed based on the existing encryption chip protocol.

[yangyao]

Do you know that this sometimes generates errors?

Sometimes is an understatement. I have seen it go wrong quite a lot, almost in every communication attempt it fails only to recover due to repeated tries.

In short, because the hardware of 1013D and 1014D is inherently bad, they have long been abandoned. A replacement version will be available in a few months this year. The same price and performance are many times higher.

[yangyao]

Just to thank everyone for their support, I don't want to sell this product anymore. I will directly replace it with the 150MHz, 1.6GS version at the same price.

[pcprogrammer]

@yangyao: A question for you. Can you tell us what the "special IC" in the 1013D, 1014D and possibly the DPOX180H is. I'm talking about the 8 pin chip that is connected to the FPGA based on a more or less I2C hardware protocol, and that "translates" information send from the F1C100s to something else based on a fixed algorithm?

And can you explain the reasoning of why this setup has been used?

Is this chip also used in any other of your designs like for instance the DPOS350P. I can not see if it is there. The 8 pin chip besides the Gowin FPGA might be it or it is the FLASH for the FPGA configuration.

A bit of advice in relation to becoming a more professional test equipment provider, is to make sure that the FPGA configuration can be updated from the MCU. For instance the Hantek DSO2000 series, that also uses a Gowin FPGA, uses the slave serial configuration option the FPGA provides. This means that the bitstream resides in the FLASH memory that also holds the firmware.

That is an encryption chip, which cannot be cracked. I blocked all information interfaces, including physical bus information, ARM IO interface, ARM algorithm, FPGA algorithm, everything is random, in order to prevent other peers from directly copying the product and causing huge losses.

Only when you changed the setup of it after developing the 1013D and 1014D then. Because in those two I cracked it with not to much effort.

If the original software is not needed and the software is newly developed, then the encryption chip is useless. However, in the new product DPOS350P, the encryption chip has been completely separated from the ARM software system, which means that the newly developed software must also be developed based on the existing encryption chip protocol.

Sounds like a challenge. 

[Algoma]

I do like that Fnirsi has been improving their products in areas of hardware and quality design. While low cost does have the largest sales there is a risky and slippery bottom when all quality becomes lost. Fnirsi gaining the reputation for products that can perform their expected duty is most valuable to bring back buyers for newer and improved products. I would certainly be interested in trying this new scope when it becomes available.

[dibu]

If the original software is not needed and the software is newly developed, then the encryption chip is useless. However, in the new product DPOS350P, the encryption chip has been completely separated from the ARM software system, which means that the newly developed software must also be developed based on the existing encryption chip protocol.

What does this mean (for the dpos250p), you can't modify or put completely new sw on it? Why would you do that? To protect your software is one thing but why limit the possibility to have different sw on it? If this locks down the device then this is a deal-breaker for me. Sorry but being hackeable is a must. Like when the old scopes were hacked did you loose any money on it? if anything it generated even more sales.

[pcprogrammer]

I do like that Fnirsi has been improving their products in areas of hardware and quality design. While low cost does have the largest sales there is a risky and slippery bottom when all quality becomes lost. Fnirsi gaining the reputation for products that can perform their expected duty is most valuable to bring back buyers for newer and improved products. I would certainly be interested in trying this new scope when it becomes available.

This still remains the question. Have you looked at the image of the board. Having so many components in the front end, does not mean that it is very good. The choice of ADC's is also questionable. You can see 4 ic's with, presumably, each two ADC's in them. Even for a single channel it will be difficult to get the signal delays and linearity properly set across the 4 ADC's to get 500MSa/s, let alone combining the two channels to get the 1GSa/s.

I can be wrong of course, but I do think that these ADC's are the same as used in the 1014D. Maybe even overclocked to do 125MSa/s per ADC, but it can also be a proper 125MSa/s device. For higher sampling rates on a single ADC more data pins would be needed to do LVDS.

If the original software is not needed and the software is newly developed, then the encryption chip is useless. However, in the new product DPOS350P, the encryption chip has been completely separated from the ARM software system, which means that the newly developed software must also be developed based on the existing encryption chip protocol.

What does this mean (for the dpos250p), you can't modify or put completely new sw on it? Why would you do that? To protect your software is one thing but why limit the possibility to have different sw on it? If this locks down the device then this is a deal-breaker for me. Sorry but being hackeable is a must. Like when the old scopes were hacked did you loose any money on it? if anything it generated even more sales.

Looking at the board gives me the impression that it will still be hack able in the sense that you can write code for the MCU and store it in the winbond flash memory (marked in green) that is located directly besides the R11 MCU. The "encryption" ic is most likely the one next to the FPGA (marked in yellow). The FPGA has it's own internal flash for the configuration.

Cracking the FPGA commands will still be a big task though. Or one has to write it's own configuration for the FPGA, but that also needs a lot of reverse engineering to get the meaning of every pin of the FPGA.

If the probably ball grid ic next to the R11 is anything special, than it might be a different story. Marked in red.

[tunk]

1014D will have a new replacement this year, with the same price, 1.6GS/s sampling rate, 150MHz bandwidth, and no false labeling

Another issue with the 1014D is the low vertical sensitivity (50mV/div).
Will this be addressed in the new model?

[yangyao]

1014D will have a new replacement this year, with the same price, 1.6GS/s sampling rate, 150MHz bandwidth, and no false labeling

Another issue with the 1014D is the low vertical sensitivity (50mV/div).
Will this be addressed in the new model?

The new model is 200MHz bandwidth, 1.6GS sampling rate, 2mV vertical sensitivity, 200M oscilloscope + 20M signal generator + 20M frequency response analyzer

[s-petersen]

If it meets it's specs this time...

[tunk]

The new model is 200MHz bandwidth, 1.6GS sampling rate, 2mV vertical sensitivity, 200M oscilloscope + 20M signal generator + 20M frequency response analyzer

And what's the memory depth?
Hopefully a bit more than the 47k in the DPOS350P.

[Sorama]

If the original software is not needed and the software is newly developed, then the encryption chip is useless. However, in the new product DPOS350P, the encryption chip has been completely separated from the ARM software system, which means that the newly developed software must also be developed based on the existing encryption chip protocol.

What does this mean (for the dpos250p), you can't modify or put completely new sw on it? Why would you do that? To protect your software is one thing but why limit the possibility to have different sw on it? If this locks down the device then this is a deal-breaker for me. Sorry but being hackeable is a must. Like when the old scopes were hacked did you loose any money on it? if anything it generated even more sales.

This is shameless.
A must??
Why don’t you go and tell Keysight, R&S, Siglent, Tek that being hackable is a must or a dealbreaker?
 

[dibu]

This is shameless.
A must??
Why don’t you go and tell Keysight, R&S, Siglent, Tek that being hackable is a must or a dealbreaker?
 

that's right, a must  , this guided my purchase decisions for quite some time now, and I don't see it change for the foreseeable future.

The difference is, is it a professional/calibrated equipment that you use in a professional setting that you can trust, or it is a toy with greatly exaggerated specs?
For sure you are not putting fnirsi in the same cathegory as the ones you mentioned.

I don't expect this from big players, that are literally using leading edge stuff, but i expect this from toys, garbage and lies don't need protection.

Many times the sole reason i bought these cheap "instruments" was the hackeability factor, and i did hack most of them, the fun, understanding and reverse engineering the firmware, plus the experience you gain from that it worth far more than the "instrument" itself, because let's face it, as is it is pretty much unusable, unreliable

[gyfhgyfh]

There is actually significant room for improvement in the 1013D and 1014D oscilloscopes. For example, adding single-channel cross-sampling at 250 MHz × 2 = 500 MHz (with an analog bandwidth of 250 MHz), or implementing sine interpolation up to 1 GHz sampling or multiple averaging (effective for repetitive signals), should be sufficient for most applications.
Under current conditions, firmware adjustments may be required to achieve 1 GHz sine interpolation sampling.