dspic30f4011 4 sample pulses

[maca404]

Is this variation designed to be used with your latest variation of the front end, I noticed it has a ground balance channel for sampling during tx has this been implemented in the pi chip ?. Also if you where to do a 4 sample scheme on chip you would only need 1 sample line which would pulse 4 times in a cycle at the appropriate times for each sample, I have not run the chip yet but it does not seem to be setup for that.

[moodz]

Is this variation designed to be used with your latest variation of the front end, I noticed it has a ground balance channel for sampling during tx has this been implemented in the pi chip ?. Also if you where to do a 4 sample scheme on chip you would only need 1 sample line which would pulse 4 times in a cycle at the appropriate times for each sample, I have not run the chip yet but it does not seem to be setup for that.

The 4 sample scheme is for a goldscan type circuit where you need 4 "parallel" samples ( demods) feeding two differential integrators ..... A single channel sampling scheme would need to route the samples and would not support overlapping sample windows. From a hardware viewpoint the 30f4012 provides three PWM outputs and two output compare OC outputs. The PWM outputs are useful for generating TX pulses however the OC outputs are particularly useful for generating sample pulses.
The 30f4011on the other hand has 4 OC outputs. So the 4 sample scheme wil only be really feasible with the 4011. Of course additional pulse timing could be generated using machine code however the the OC unit will run without needing hand tuned code loops.
The ground sample is experimental and based on the change in peak coil current caused by ground effect during the tx pulse. The current released hex code does not support this yet ... I have only placed it on the schematic for comment.
Moodz

[maca404]

Very interesting I would be interested to see what information can be had from the ground channel. With your current iteration of the DIF front end could you give a short run down on how this front end interacts with the PI chip as in samples and filters etc ?

[IBGold]

[moodz]

Very interesting I would be interested to see what information can be had from the ground channel. With your current iteration of the DIF front end could you give a short run down on how this front end interacts with the PI chip as in samples and filters etc ?

Yes it would be interesting .. however I have not done done the field testing so that is up to the dear readers .... I just supply the ideas. However having said that there is an observable effect on peak coil current during the Tx pulse due to the ground ... so surely ?? there must be a useable signal here for ground balance ( after all ... the biggest target under the coil ??? ... is the ground itself. ) The UNIPI chip is fairly simple ... it merely provides a fully adjustable TX pulse and multiple fully adjustable sample pulses that are crystal locked. The analogue portion only applies a integration and time differential detection algorithm to an incoming voltage on the analogue input. I would expect that some analogue processing and filtering was done prior .... the digital part will compensate for drift by the analogue bits.

moodz

[maca404]

please excuse this being slightly off topic, I just got some 30f4011 chips and a clone pickit3 and the icsp breakout board with the zif sockets. Whenever I try and read/write the chip I get "unknown device detected", remove the pic I get "no device" and when I turn the power on and off it all work so I can only assume the pickit3 is ok. I think it may be a osc issue I tried setting the config for internal osc but still no read/write, can someone tell me if I have to have an external oscillator for the initial program of the chip or does the pickit3 produce its own clock for programing?.

[Dr Vel]

Excessive ribbon length is a minor possibility and the major likelihood is you have insufficient USB power to your Pickit3, especially if you are using a USB hub. Other things could be the cause but the USB power issue is an important parameter for reliable programming. I have had this issue with other model USB programmers in the past (K149 and K150 both of which do that with poor USB ports). I imagine an externally powered hub would be OK depending upon how well designed it's power supply is, but I now prefer to never use a hub with a programmer. I have not yet used this number and have not looked it up but be sure the Pickit3 covers it. I am pretty sure it should, and of course other things could be going on in your case. However never overlook the simple reason, especially if your USB port is coming from a notebook rather than a desktop computer.

[moodz]

please excuse this being slightly off topic, I just got some 30f4011 chips and a clone pickit3 and the icsp breakout board with the zif sockets. Whenever I try and read/write the chip I get "unknown device detected", remove the pic I get "no device" and when I turn the power on and off it all work so I can only assume the pickit3 is ok. I think it may be a osc issue I tried setting the config for internal osc but still no read/write, can someone tell me if I have to have an external oscillator for the initial program of the chip or does the pickit3 produce its own clock for programing?.

Oh oh ..! the clone pickit3s don't work so good. Power the chip independently making sure all ground and vcc are connected ( there is more than one ) tie MCLR pin to +5 using 10k resistor.
For programming the device the clock setting should not matter.

[maca404]

Thanks, I did as you said and now I actually got it to say its programmed. However it fails to verify, interestingly If I now read the chip it looks exactly like the .hex except the first 2 bits which read 0x00 ?. Weird

[moodz]

Thanks, I did as you said and now I actually got it to say its programmed. However it fails to verify, interestingly If I now read the chip it looks exactly like the .hex except the first 2 bits which read 0x00 ?. Weird

Does it say something like "fails verification at 0x0000 expected 0x72 read 0x00 or something similiar ?? the code protect bits might be set to on ... so it wont read back.

moodz

[maca404]

Yes the error is like that, I assume it may be that as well but I can read the chip and it is only the first two bytes that differ from the original hex. Plus I can erase the chip and then it still wont verify blank however the whole chip will read 0xfff except the first two bits which again read 0x00. I also tried a second chip and same thing.

[moodz]

Yes the error is like that, I assume it may be that as well but I can read the chip and it is only the first two bytes that differ from the original hex. Plus I can erase the chip and then it still wont verify blank however the whole chip will read 0xfff except the first two bits which again read 0x00. I also tried a second chip and same thing.

two bytes or two bits ?? .... just to be clear.

[maca404]

two bytes or two bits ?? .... just to be clear.

Bytes lol

[moodz]

Oh oh ..! the clone pickit3s don't work so good. Power the chip independently making sure all ground and vcc are connected ( there is more than one ) tie MCLR pin to +5 using 10k resistor.
For programming the device the clock setting should not matter.

My web research indicates that there is a known bug where the timing on the MCLR pin is not that good result in a first or second byte fail to verify or blank check. However programming should be good. If you have a CRO with storage see if you can capture the pulse on MCLR pin when verify or blank check starts up vs the data lines. The voltage level on MCLR pin is important too.

[maca404]

I did put the scope on it today and it seemed ok and there was no noise but I did not go in to much detail, Interestingly if I look in detail at what is read vs the actual hex there is some minor errors so I am wondering if this could be a noise issue or something. ultimately I may have to try another programer this clone could be a dud.