Single capacitor integrator switch placement

[green]

I had remembered the fridge magnet giving a signal. Tried this morning with target response tester, scoped amplifier out no integrator. 8inch figure8 coil. Couldn't see change with fridge magnet. Ceramic hobby magnet easy to see signal as I pulled magnet away from coil. Neodymium magnets gave a similar response but they are a short TC targets static. Must have done something wrong when I tried adjusting EF with the fridge magnet before.

[dbanner]

In the MPP integrator, the input is coupled to the output during the off period via feedback resistor. Don't know if it matters in this application. Pulling the input before the fet to ground may have some advantages. Not sure.

[green]

Got a packet of eight 20mm ceramic magnet's, quite strong for the size.

We used ceramic magnets taped to prop shafts with hall effect pickup for recording prop speed in dusty environments. Purchased a pack of magnets that didn't work. Had N and S on same side, if I remember right they were quite strong. Don't remember if we could tell by looking at them, could easily see if looking at signal on a scope as prop rotated.

Not sure it matters if pulling magnet away from center of coil causes a signal when adjusting for EF and magnet doesn't cause a signal change when placed on coil.

[green]

: I ran a simulation whereby the lines were pulled to ground during the off periods. V(out) looking much the same. I suppose more dynamic testing is needed.

If VIN goes enough negative, the fets would turn on. Pulling the lines to ground should prevent them from turning on. Other wise not much difference.

[6666]

ceramic magnet's

Got a packet of eight 20mm ceramic magnet's, quite strong for the size.

[dbanner]

: I ran a simulation whereby the lines were pulled to ground during the off periods. V(out) looking much the same. I suppose more dynamic testing is needed.

[dbanner]

Carl will have to pull out the heavy artillery from his bag of tricks and give us long awaited analysis of the MPP integrator.

[dbanner]

Pulling the lines before the fets to ground is bound to change things, but I'm not sure what will be the effect, especially with respect to the intended purpose of the circuit collaboration with the sample switches and it's intended functionality.

[bklein]

Carl, I think it may be in one of Electronics magazine?s case books.

[dbanner]

Thanks, first two are for normal integrator. Last is MPP integrator with a constant input. Same result as I got with constant input. Played with spice some more. Got what I thought simulated target decay during sample. Increasing sample time caused the signal to be less where signal increases for longer TC targets when increasing sample time with a 1C integrator . Maybe I'm doing something wrong?

Changing sample time or sample rate changes integrator TC with the track hold differential integrator but not with 1C integrator. Maybe not a problem?

Not so easy is this MPP integrator. I have only vague understanding. I suspect having the feedback resistor disconnected from the input when the fet is off makes huge difference to C(hold).

A fixed Vin with various sample widths won't show up the true nature of this beast. I couldn't imagine doing a dynamic analysis using old school slide rules and pen and paper, much respect to the engineers of old.

Young players have it so easy with powerful simulation software nowadays.

[6666]

EDN has the Design Ideas archived back to 1969.



If you're talking about a flexible strip fridge magnet, those are Halbach magnets which have a low flux density except near the surface. Ideally, one surface has an alternating +-+-+- field and the other side has no magnetic field at all. I would avoid this for testing EFE. Find a good ceramic magnet that doesn't create any kind of target signal.

Yes talking about the strip magnets, I was not convinced my testing was working properly , so I will get some ceramic magnet's and try, thanks for the tip.

[green]

Green here are some threads I dug up. We all have read these before, but may spark some new understandings

See post #4

https://www.geotech1.com/forums/showthread.php?13113-Designing-an-Integrator-circuit&highlight=integrator+gain

Calculating integrator gain

https://www.geotech1.com/forums/showthread.php?19079-Calculating-integrator-gain&highlight=calculating+integrator+gain

Simulation

https://www.geotech1.com/forums/show...ntegrator+gain

Thanks, first two are for normal integrator. Last is MPP integrator with a constant input. Same result as I got with constant input. Played with spice some more. Got what I thought simulated target decay during sample. Increasing sample time caused the signal to be less where signal increases for longer TC targets when increasing sample time with a 1C integrator . Maybe I'm doing something wrong?

Changing sample time or sample rate changes integrator TC with the track hold differential integrator but not with 1C integrator. Maybe not a problem?

[Ferric Toes]

https://www.geotech1.com/forums/atta...4&d=1591477213

I've spent a lot of time trying to understand the track hold differential integrator without much success. Good because it keeps the brain active, one of my reasons for being on the forum.

Think I understand a 1C integrator(output=Rfdbk/Rin*sample time*sample rate*average sample V) TC=(Rfdbk*Cfdbk). Sample rate and sample time effect output signal.

Using spice with constant 1V input with track hold differential integrator, output=1*6800/2200*((Rfdbk/(Rfdbk+10k))second stage. Doesn't appear sample time or rate effect gain with constant input. Wondering how sample time effects signal out when sampling a quarter? Don't know how to calculate TC.

Here is a page from an Analog Devices tutorial MT-090 that gives the basic idea. It also shows a second switch from line to ground as dbanner had a question about in the nickel/quarter thread. I think I have more info and will look it up tomorrow.



Eric.

[Altra]

https://www.geotech1.com/forums/atta...4&d=1591477213

I've spent a lot of time trying to understand the track hold differential integrator without much success. Good because it keeps the brain active, one of my reasons for being on the forum.

Think I understand a 1C integrator(output=Rfdbk/Rin*sample time*sample rate*average sample V) TC=(Rfdbk*Cfdbk). Sample rate and sample time effect output signal.

Using spice with constant 1V input with track hold differential integrator, output=1*6800/2200*((Rfdbk/(Rfdbk+10k))second stage. Doesn't appear sample time or rate effect gain with constant input. Wondering how sample time effects signal out when sampling a quarter? Don't know how to calculate TC.

Green here are some threads I dug up. We all have read these before, but may spark some new understandings

See post #4

https://www.geotech1.com/forums/showthread.php?13113-Designing-an-Integrator-circuit&highlight=integrator+gain

Calculating integrator gain

https://www.geotech1.com/forums/showthread.php?19079-Calculating-integrator-gain&highlight=calculating+integrator+gain

Simulation

https://www.geotech1.com/forums/showthread.php?23628-MPP-Sampling-Integrator-Simulation&highlight=integrator+gain

[green]

If you're talking about a flexible strip fridge magnet, those are Halbach magnets which have a low flux density except near the surface. Ideally, one surface has an alternating +-+-+- field and the other side has no magnetic field at all. I would avoid this for testing EFE. Find a good ceramic magnet that doesn't create any kind of target signal.

Round ceramic magnets can have +- on the same side also. Need to be more selective than I was thinking.