Bandido II - more stuff

[Max]

Hi all,
I've found a little "problem" around manual setting in bandido. I've built one and have good results now but find a little problem that is:
- in manual operation threshold slowly vary thus giving a reduction of depth

I know that manual setting (non-motion) is related to pinpointing (read the manual)...but noticed that is accurate and sensitive for say 1 minute then require retune...

So seems not a real problem...well, if one like to use as searching mode too ? having a non-motion behaviour ? I like non-motion detection but know that this way a d.c. coupled op-amp could drift a lot...

I think that auto-mode and disc are good...motion type...detection but sometimes you need to have a great speed between targets (small) and coils to detect them (>1m/s). And that's not so good.

Question are:
1. non-motion: is possible to reduce threshold drift in manual-mode (all-metal)?
2. motion: is possible to reduce speed relative in auto (all-metal) and disc mode ?

Any idea ? Anyone experienced too same behaviours/wishes ?

Best regards,
Max

[Delbert grady]

get my drift

Dear Max, non motion detectors always drift. If you can keep the threshold for a minute thats quite good. Thats why the motion detector was invented.

Delbert

[Max]

Dear Max, non motion detectors always drift. If you can keep the threshold for a minute thats quite good. Thats why the motion detector was invented.

Delbert

Hi Delbert,
yes I know...I've used many other non-motions before, always drift. But question is if is it possible to reduce drift speed here.
Some years ago my cousin gives to me his non-motion detector and I realized it drifted a lot and was almost unstable. I realized it was due to fet stage and minority carriers involved due to thermal drift.
Using tubes I've redesigned osc. stage and then it worked ok , retune needed every 20-30 minutes where before any 30-60second or less you have to move threshold pot...
And my cousin still thanks when we talk about that detector. But in other cases I haven't similar success...so often gived up and maked motion mod on these - and then no-drift at all.

Anyway, I don't know if it's possible in Bandido and how...just asked if someone have already studied the issue or a simple mod.
Also retune speed is a big issue in motion modes I think...here on bandido circuit - something that can be modified easy say changing C37 value.

Best regards,
Max

[chemelec]

Hi Delbert,
yes I know...I've used many other non-motions before, always drift. But question is if is it possible to reduce drift speed here.
Some years ago my cousin gives to me his non-motion detector and I realized it drifted a lot and was almost unstable. I realized it was due to fet stage and minority carriers involved due to thermal drift.
Using tubes I've redesigned osc. stage and then it worked ok , retune needed every 20-30 minutes where before any 30-60second or less you have to move threshold pot...
And my cousin still thanks when we talk about that detector. But in other cases I haven't similar success...so often gived up and maked motion mod on these - and then no-drift at all.

Anyway, I don't know if it's possible in Bandido and how...just asked if someone have already studied the issue or a simple mod.
Also retune speed is a big issue in motion modes I think...here on bandido circuit - something that can be modified easy say changing C37 value.

Best regards,
Max

Not Sure, But would the stability not have more to do with the Storage Quality of C35?

[Max]

Not Sure, But would the stability not have more to do with the Storage Quality of C35?

Hi Gary,
yes is true. Also for me is also involved capacitor leak ratio and dielectric histeresys. I've tried with polistirene and polyester at now...no big differences.
My idea is: in motion mode IC6a acts as sample/hold (actually hold only when in manual - sample when you retune...and also as feedback block to IC6b).
But drop speed here is above >30uV/s much more that I want...so capacitor is involved for sure...but also IC6 bias current...IC6 input resistance (one branch)... IC6 differential resistance...etc etc

We end-up with a non-infinite sample-and-hold here...and it drifts a lot after 1 minute so manual mode is well, this way, only in pinpointing.

My idea is alter the schema a bit... but just want to know if someone tried this way before ...and eventual results if any.

I know that in original bandido manual-mode (all-metal) is only for pinpointing...thus require 30-60 seconds( maybe less)...not more, so it's not a "problem" or "malfunction" of original...just my curiosity.

Best regards,
Max