1. Home
  2. Electronics Design
  3. There are Idiots and There are Idiots

There are Idiots and There are Idiots

This design group I am in is being led by a mechanical engineer who has his limitation. Initially I thought it was just that he didn't know much abou t electronics. Then I found his thinking for mechanical issues to be a bit lacking in sophistication. Now I realize he doesn't understand fundamenta l principles of analysis.

I was not involved in the project from the start. They had built a revisio n 1 of the board which was essentially an arduino with a motor controller c hip and some comparators. I realized they were not very skilled at electro nics design and figured I could help.

What I didn't realize was the extent of ignorance in project development an d management. The entire project has moved along in fits and spurts and ma ny, many backflips because of a lack of initial analysis and specification of requirements.

All through this project they have been planning to measure flow rate by me ans of detecting the differential pressure across an orifice. They had a g uy who seemed to have that as his sole task so I expected it was covered. There were concerns about the resolution at the low end so a separate circu it was added to provide more gain for the low end giving more resolution. Again, I expected they had it under control.

I am designing much of the FPGA that reads the differential sensor and othe r inputs to calculate the flow rate. The ADC will be a delta-sigma built i n the FPGA with over 17 bits of resolution. I'm thinking that is enough, b ut to be sure I ask for a number for the resolution in flow rate required.. .

No one could provide that simple piece of data. It has gone on for weeks w ith every excuse in the book for not providing it. Instead they talk about having to take into account all manner of errors and could not explain wha t they were talking about. Today I finally found out they want to determin e the resolution based on it's contribution to accuracy. Once I had that d ata it took me about 15 minutes to do the calculations that show even if th ey assume no other source of error the resolution required would be about 1 part in 15,000 or 14 bits of ADC. That's great, but this worst case happe ns at the low end of the scale and the error of the pressure sensor itself is pegged to the full scale value and so is enormously large at this close to zero.

I think they are now trying to rationalize continuing down this road by pic king a higher value as the minimum to be measured. Trouble is they don't h ave control over that.

Months ago I suggested they go with a flow sensor that was cheap, available and produced a digital output of 14 bits, adequate for this purpose. That was rejected because they were new and in short supply. However, the pres sure sensors used in ventilators are all in short supply. So it doesn't ma tter much which device you pick if they are all in short supply.

I'm about fed up with this project. We can replace anyone on the project, except for the guy running it, the one we need to replace. Sounds like man y of the jobs I've had.

--

Rick C. 

- Get 1,000 miles of free Supercharging 
- Tesla referral code - https://ts.la/richard11209
Reply to
Rick C
Loading thread data ...

tirsdag den 8. december 2020 kl. 00.03.08 UTC+1 skrev snipped-for-privacy@gmail.c om:

her inputs to calculate the flow rate. The ADC will be a delta-sigma built in the FPGA with over 17 bits of resolution. I'm thinking that is enough, b ut to be sure I ask for a number for the resolution in flow rate required.. .

I'd be skeptical about how much those 17bits are worth, most MCUs with buil d ADCs can bare do 10-12bits

Reply to
Lasse Langwadt Christensen

.com:

other inputs to calculate the flow rate. The ADC will be a delta-sigma buil t in the FPGA with over 17 bits of resolution. I'm thinking that is enough, but to be sure I ask for a number for the resolution in flow rate required ...

ild ADCs can bare do 10-12bits

The more I learn about the FPGA ADC the more I like them. The FPGA only co ntains the comparator (on a separate Vcc from the rest of the chip) and the digital logic. I took the precaution of using an external CMOS driver for driving the analog portion of the ADC which is all external to the FPGA.

I dug around quite a bit and I never found a work that actually shows the l imitations of these ADCs in the context we will be using them. Unfortunate ly no one actually did an error analysis on the flow rate sensor until I st arted being a PITA about the required resolution. This eventually showed w e were not able to meet the accuracy spec with the device we had picked mon ths and months ago. In fact, this episode might be the straw that breaks t he camel's back. I really can't continue working when each couple of weeks we find something significant that isn't working.

I would like to see the result of the 17.3 bit ADC and I would like to comp lete the calculator design in the FPGA. But what's the point when there's a 50/50 chance of tossing it all to start over on a new approach.

--

Rick C. 

+ Get 1,000 miles of free Supercharging 
+ Tesla referral code - https://ts.la/richard11209
Reply to
Rick C

As far as I know (which is not much), the main problem of ADCs implemented in this way is not the resolution, but rather the absolute accuracy.

I suppose that for the comparator you are using the inputs normally used for LVDS/LVPECL. Even if you use a "clean" Vref, your feedback will run through digital I/O buffers, and therefore also the Vccio supply will contribute to the system accuracy.

You will need an insanely accurate power supply in order to get 17 bits, we are talking microvolts here. Even if you achieve it, the (very small) fluctuations due to the I/O switching will slightly change the output voltage of the buffers.

I'm quite interested, let us know how it goes if your project leader does not screw up everything. :)

--
Fletto i muscoli e sono nel vuoto.
Reply to
dalai lamah

d

we

Yes, accuracy is hard to obtain. I had the impression they had done a prop er analysis of accuracy, precision and resolution.

However, the ADC will have characteristics that can be compensated for. I' ve done an initial analysis of the errors and most will be correctable with a scale factor and offset which is done for the equivalent errors of each sensor. The remaining error will be a curve due to charge injection, large st in the center of the range and approaching zero at the scale ends. Ther e is also error from the accuracy of the two resistors which can be minimiz ed, but not eliminated due to temperature drift. The important factor is t he ratio, so mounting them so the temperature variation is minimized helps.

I'm still processing the realization of how bad the leadership is, trying t o determine if it is worth continuing to contribute to the development. He did manage one thing... he has found a company in Brazil who is interested in considering our machine for productization. I suggested that rather th an wait for our efforts to be completed (and making any number of decisions they will end up changing) that we allow them to review our project now an d have some input to the decisions we are making. In spite of the fact tha t everything we are doing is licensed, open source materials he feels getti ng any feedback from this company would open us to liability.

I need to knock out the calculator code. Once I have something that simula tes I can leave the project and not feel that I've left them in the lurch. They can maintain it as they continue to reinvent the same designs over an d over.

Reply to
Rick C

Right. Delta-sigma has continuously squiriming numbers of rising and falling edges, at high rates, and any rise/fall rate differences become errors. You can express, say, a 41% duty cycle lots of different ways in a serial data stream.

Or kill a lot of people.

--

John Larkin      Highland Technology, Inc 

The best designs are necessarily accidental.
Reply to
jlarkin

In floating point normalize the value so that the exponent is even i.e. the mantissa is between 1 .. 4. Use a polynomial for the mantissa to calculate the square root, which will be between 1 .. 2.. A 3rd or

4th order polynomial should be enough for single precision accuracy. Divide exponent by 2 and combine it with the polynomial result.
Reply to
upsidedown

d

we

I don't think I mentioned that the power supply doesn't need to be good to microvolts because the sensors are ratiometric to the supply. If the suppl y goes up the signal from the sensor goes up in kind. So the ADC reference needs to go up as well to compensate. So the issue here is not a hugely a ccurate fixed reference, but rather keeping noise out so the supply voltage is the same all around. Then the 17.3 bits is obtained by integration of the samples for 5 ms, which is a pretty awesome filter.

Not at all trying to get 17 bits of accuracy really, just good accuracy and good resolution, meaning something like an MCU provides. The integration period is set by the update rate so we get 17.3 bits of counting if we like it or not. The math logic is in 18 bit hunks, so this matches pretty well .

--

Rick C. 

--+ Get 1,000 miles of free Supercharging 
--+ Tesla referral code - https://ts.la/richard11209
Reply to
Rick C

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.